Hastings Constitutional Law Quarterly Volume 45 Article 4 Number 3 Spring 2018

1-1-2018 Striking the Right Balance: Hate Speech Laws in Japan, the United States, and Canada Craig Martin

Follow this and additional works at: https://repository.uchastings.edu/ hastings_constitutional_law_quaterly Part of the Constitutional Law Commons

Recommended Citation Craig Martin, Striking the Right Balance: Hate Speech Laws in Japan, the United States, and Canada, 45 Hastings Const. L.Q. 455 (2018). Available at: https://repository.uchastings.edu/hastings_constitutional_law_quaterly/vol45/iss3/4

This Symposium is brought to you for free and open access by the Law Journals at UC Hastings Scholarship Repository. It has been accepted for inclusion in Hastings Constitutional Law Quarterly by an authorized editor of UC Hastings Scholarship Repository. For more information, please contact [email protected]. Striking the Right Balance: Hate Speech Laws in Japan, the United States, and Canada

by CRAIG MARTIN*

Introduction The issue of hate speech has occupied the headlines of many democracies in recent years. Whether it be Charlottesville, Kyoto, or Warsaw, the rise of nativist, nationalist, and racist groups, expressing hatred towards minorities within society, has once again confronted us with the question of how far democracies can or should go in limiting certain extreme forms of hateful discriminatory expression. Certain types of hate speech, which have the purpose and effect of fostering hatred against groups defined by certain characteristics such as race, ethnicity, religion, gender, or sexual orientation, and which take the form of extreme vilification, denigration, and even dehumanization of its targets, are known to cause significant harm.1 This harm is suffered by the members of such groups, both in the form of direct emotional and psychological harm caused by the speech itself, and also in the form of increased levels of discrimination and persecution as a consequence of the hatred fomented in society by such speech. These increased levels of discrimination also do significant harm to the broader constitutional system, by undermining the democratic principles of equality

* Professor, WashburnUniversity School ofLaw, B.A. (R.M.C.), J.D. (Univ. of Toronto), LL.M. (Osaka Univ.), S.J.D. (Univ. of Pennsylvania). This article was developed out of a much shorter chapter, {t H * 0) & Laws Without Sanctions:Hate Sp eech Laws and the BalancingofRights in Japan, in 1 W [THE LEGAL PROCESS IN CONTEMPORARY JAPAN: ITS STRUCTURE AND DYNAMICS] (Keiichi Ageishi et al. eds., 2017), substantially expanded to provide more robust comparative and theoretical analysis of the tension between rights posed by hate speech laws. I would like to thank Setsuo Miyazawa and Keith Hand for inviting me to present this at a special symposium on Japanese hate speech law at UC Hastings College of the Law, and the other participants of that symposium for their feedback and comments, in particular Jeff Adachi, Hiroshi Fukurai, Junko Kotani, Rory Little, David Makman, Setsuo Miyazawa, and Alice Yang. In addition, I would like to thank, for their comments and input at various stages of the project, Katie Baylie, Lois Chiang, Benson Cowan, Ayako Hatano, Jeff Jackson, Ali Khan, Mark Levin, Bill Rich, Takashi Shirouzu, and Frank Upham. 1. See infra notes 38 and 174-181 and accompanying text.

455 456 HASTINGS CONSTITUTIONAL LAW QUARTERLY [Vol. 45:3 and inclusiveness, and ironically by undermining freedom of expression itself, as the members of target groups are cowed into silence.2 Efforts to prevent such harm through the imposition of legal limits on this form of hate speech, however, would likely impinge upon the constitutional right to freedom of expression. Freedom of speech is viewed as a fundamental individual right in most democratic constitutional systems and is a right that international human rights law obliges states to guarantee, respect, and enforce. It is difficult to conceive of an approach to the prohibition of hate speech that would not constitute a government suppression of political expression based on its content or even its viewpoint, not just its form or its effects. Such content-based limitation on political speech is viewed with the greatest suspicion, and in the context of judicial review, the justification of this kind of limitation on the constitutional right of freedom of expression is difficult (though by no means impossible). This is because the regulation of political speech is seen as being inimical to the most fundamental values at the core of the right to freedom of expression, and as doing considerable violence to principles thought essential to the operation of a free and democratic society. We are thus confronted with two threats or potential harms to important values. If we take the harm of hate speech seriously, we must accept that it poses a threat to democratic values, and as will be argued here, even undermines constitutional rights. But to suppress hate speech similarly risks causing significant harm to constitutional rights and democratic principles. The challenge faced by constitutional democracies, therefore, is determining how to reconcile these competing imperatives, how to resolve this tension. Where to draw the line, how to find the right balance? Upon further reflection, and an examination of the approaches in different countries, we might further ask what, exactly, are we even trying to balance? For a comparative review of different approaches to the problem of hate speech suggests that different countries are trying to balance different things. This Article uses the occasion of Japan's enactment of hate speech legislation in 2016, the latest effort by a democratic system to grapple with the problem, to engage in a comparative examination of three very different approaches to resolving the tensions outlined above. The Article explores the purpose and scope of hate speech related laws, and the manner in which such laws have been judicially reviewed, in Japan, the United States, and Canada. It does so with a view to better understanding the different ways in which the balance might be struck and assessing whether there may be an optimal approach to achieving some equilibrium. There are significant

2. See infra notes 38 and 174-181 and accompanying text. Spring 2018] STRIKING THE RIGHT BALANCE 457 differences in all aspects of the approaches of the three systems. To begin with the legislative approach, the purported object and purpose, the mischief targeted, and the breadth and scope of the expression contemplated by hate speech related laws, are all quite different. With respect to the judicial approaches, how such law has been characterized and interpreted by respective judiciaries, and the constitutional doctrine that the law is subject to in judicial review, is also strikingly different. Identifying and reflecting upon some of these marked differences can assist in thinking about what formulation of hate speech law might be justified as a reasonable limitation on freedom of expression, as well as what adjustment to constitutional doctrine might be required to make such justification formally possible. There is, however, a singular and more fundamental difference that emerges from an examination of the Japanese and American approaches through the lens provided by the Canadian example. This difference is in the extent to which hate speech laws are understood to relate to the constitutional right to equal treatment and not to be discriminated against. As we will see, the Canadian legislatures and courts both quite explicitly recognize that the purpose of hate speech law is to advance the equal protection and equal benefit of law that are provided for in Section 15 in the Canadian Charter of Rights and Freedoms.3 It is precisely for this reason that the overriding concern, and the explicit object and purpose of hate speech laws, is to prevent discrimination. The Canadian system thus tends to view the justification of hate speech laws in terms of striking a compromise or balance between the constitutional rights to equality and to freedom of expression. This is very much in line with the manner in which international human rights law, as reflected in the International Covenant on Civil and Political Rights ("ICCPR") 4 and the International Convention on the Elimination of All Forms of Racial Discrimination ("CERD"),5 quite explicitly requires states to both implement hate speech laws for purposes of furthering the right to equality, and to nonetheless respect and enforce the right to freedom of expression. This is in marked contrast to both the Japanese and American approaches, which do not tend to recognize that hate speech laws implicate in any way the constitutional provisions providing for equal protection and the right not to be discriminated against.

3. Canadian Charter of Rights and Freedoms, Part I of the Constitution Act, 1982, being Sched. B to the Canada Act 1982, ch. 11, section 15(1) (U.K.) [hereinafter Charter of Rights and Freedoms]. 4. International Covenant on Civil and Political Rights, article 20(2), Dec. 19, 1966, 999 U.N.T.S. 171 [hereinafter ICCPR] (entry into force Mar. 23, 1976). 5. International Convention on the Elimination of Racial Discrimination, 660 U.N.T.S. 195 [hereinafter CERD] (entry into force Jan. 4, 1969). 4589 HASTINGS CONSTITUTIONAL LAW QUARTERLY [Vol. 45:3 It is in this sense, then, that we find that the different approaches to hate speech law are trying to balance different things. In the Japanese and American approaches, the balance is between the fundamental constitutional right to freedom of expression on the one hand, and mere legislation on the other. Put slightly differently, it is a balance between harms to constitutional values and principles on the one hand, and on the other hand sociological harms that hate speech legislation seeks to address, but which are not recognized by the courts as being very grave. Not surprisingly, the line tends to get drawn in favor of the constitutional right to freedom of expression. In the Canadian approach, however, the balance to be struck is explicitly recognized as being one between competing fundamental constitutional rights, and the harms at issue on both sides of the equation implicate and threaten constitutional values and principles. As will be explored below, Canada does not necessarily draw the line in the right place in terms of how far to limit expression, but I would argue that it is at least striking the right balance-in the sense that it is at least balancing the right things and recognizing that the tension created by hate speech law is between two constitutional rights. The argument that the "right balance" requires or constitutes a recognition that hate speech law implicates the constitutional right to equal treatment and protection, thereby creating a tension between two constitutional rights, finds support from certain features that are common to the doctrine governing constitutional judicial review across most systems. That is to say, the analytical framework of the doctrine is such that recognition of hate speech as implicating equality rights naturally gives rise to an internally coherent approach that both minimizes the law's impact on freedom of expression and shapes it in a manner that maximizes its probability of surviving judicial review across many systems-even that of the United States. If hate speech laws are understood to fulfill the purpose of the right to equal protection, there comes into view not only a constitutionally informed objective and purpose, which will obviously serve as a compelling state interest, but also a ready- made set of limiting principles that help to more precisely define the very narrow scope and tailored limits for the legislation. These internal limits will be explicitly informed by the constitutionally prohibited grounds of discrimination. The law that arises from this approach provides a much stronger basis for justifying as reasonable the resulting infringement of freedom of expression. The comparative perspective helps to provide answers to many of the concerns that are typically raised in respect of the breadth of hate speech laws and the injury they do to freedom of speech. Ultimately, I will suggest that hate speech laws along such lines, that are drawn as narrowly Spring 2018] STRIKING THE RIGHT BALANCE 459 as the Canadian CriminalCode provision 6 (but perhaps not as broadly as some of the Canadian human rights code provisions), can indeed be upheld as a reasonable and justifiable limitation on the right to freedom of expression. Before launching into the comparative examination, however, we should pause to consider what, precisely, we mean by "hate speech," and thus what should be prohibited or constrained by so-called hate speech laws, subject to freedom of speech considerations. The answer to this question is, of course, dependent on some of the other aspects of the problem with which we will grapple below, and so to answer that question here may in some ways seem to predetermine the normative arguments to follow. But it is nonetheless important to provide some preliminary observations on this point. As explained in more detail below, the Canadian system characterizes hate speech as expression that, in form and content, will likely foster hatred and contempt for individuals who are identified as belonging to groups based on such shared personal characteristics as race, ethnicity, creed, religion, or gender ("identifiable groups"), thereby causing not only emotional and psychological harm to them, but also making discrimination and persecution of the targeted identifiable groups more likely.7 This is consistent with the approach reflected in international human rights law.' On this view, the more extreme kind of hate speech that should be of concern in democratic societies, is that which vilifies, denigrates, alienates, delegitimizes and even attempts to dehumanize the members of identifiable groups-groups that often have been historically the subject of negative stereotyping, prejudice, and discrimination. And of gravest concern, and therefore the subject of possible legal limitation and constraint, is such expression that takes the most extreme form, through the use of epithets, historic symbols of hatred and terror such as the Swastika or a burning cross, and the use of dehumanizing animal metaphors for members of the group.9 This Article proceeds in four parts. The first three that follow this introduction examine the approaches of Japan, the United States, and Canada, and I should note that they do so on the assumption that readers will likely be well acquainted with one or perhaps two of these systems, but not all three-they are all, therefore, written in part for those who may not know the system well. Part I provides an examination of the new hate speech law regime in Japan, analyzing the new legislation, why it was developed, and how it should be understood within the context of Japan's historic approach

6. See infra notes 38 and 174-181 and accompanying text. 7. See infra notes 38 and 174-181 and accompanying text. 8. See infra notes 47-52, 185, and 229-231 and accompanying text. 9. See infra notes 174-181, 200-207 and accompanying text. 460 HASTINGS CONSTITUTIONAL LAW QUARTERLY [Vol. 45:3 to the constitutional rights to equality and freedom of speech. In Part II the Article examines how the American system has grappled with hate speech, exploring in particular how judicial decisions have considered hate speech laws within different and ill-fitting doctrinal categories of lesser-protected speech as a means of assessing their constitutionality. The Article turns in Part III to examine the Canadian approach, in terms of both the criminal law and human rights code provisions limiting hate speech. In this part I explain how, in contrast to the Japanese and American approaches, both the legislatures and the courts have recognized explicitly that hate speech law has the object and purpose of fulfilling the guarantee of equal protection provided for in the Charter of Rights and Freedoms. Finally, in Part IV, I explore in more detail the lessons to be drawn from this comparative review, beginning with an explanation of why hate speech law should be recognized as implicating the constitutional right to equal protection, and how such a recognition leads to laws that are naturally limited by constitutional principles, and more easily justified as a reasonable impingement on the right to freedom of expression. The comparative analysis suggests the right balance can be struck by recognizing the tension as between two fundamental constitutional rights or values and taking seriously the harm that can be caused by hate speech.

I. The New Hate Speech Law Regime in Japan We begin the comparative review with a discussion of the recent introduction of hate speech legislation in Japan and analyze how that fits into a broader constitutional approach to equality rights and freedom of speech. First, however, that legislation must be put into some perspective by explaining the developments that led a reluctant government to enact this hate speech law, and by examining the legal remedies available for addressing hate speech prior to the legislation.

A. The Call for Hate Speech Laws in Japan There have been calls in Japan for human rights legislation and antidiscrimination protections for many decades now.' 0 But the more recent

10. I also explore these issues, in the context of a comparative examination of constitutional equality rights in Japan, in Craig Martin, Glimmers of Hope: The Evolution of Equality Rights Doctrine in Japanese Courtsfrom a Comparative Perspective, 20 DUKE J. COMP. & INT'L L. 167, 176-78 (2010); for these issues in the context of hate speech, see Shigenori Matsui, The Challenge to Multiculturalism:Hate Speech Ban in Japan, 49 UBC L. REv. 427 (2016) [hereinafter Matsui, Hate Speech]; for an excellent and detailed analysis of social justice litigation, though now somewhat dated, see FRANK UPHAM, LAW AND SOCIAL CHANGE IN POSTWAR JAPAN (1987); and see also Timothy Webster, Insular Minorities: InternationalLaw's Challenge to Japan'sEthnic Homogeneity, 36 N.C. J. INT'L & CoMM. REG. 557 (2011). Spring 2018] STRIKING THE RIGHT BALANCE 461 and earnest appeals for the enactment of hate speech laws began in 2012, in response to an increase in the incidence of anti-Korean rallies and demonstrations. There are over half a million people in Japan who are of Korean descent and who do not have Japanese nationality. They are mostly descendants of immigrants from Korea who were stripped of their Japanese nationality after World War II. While they are third and fourth generation descendants of those initially brought to Japan when Korea was part of the Japanese Empire, and who were thus born and raised in Japan, they retain either North or South Korean nationality, depending on where their forefathers came from, and have a special permanent resident status (similar to descendants of Taiwanese in Japan, who were similarly stripped of their nationality after the war). Indeed, some are effectively stateless as a matter of international law, even though registered as being North Korean under Japanese law." The internal dynamics of this community are complex, and those of Northern and Southern Korean identity each suffer different forms and degrees of discrimination, but I will here refer to them collectively as Korean-Japanese.1 2 Several of the rallies against Korean-Japanese were videotaped and received considerable publicity, including one in particular which featured a young girl screaming that Koreans should be massacred, among other things.1 3 The protests were characterized by speeches and chants that included highly discriminatory and hateful invective aimed at Korean-Japanese persons. Several of the protests were held outside of Korean-Japanese schools, and so subjected young children to abusive verbal attacks. The first ever government study of the issue in 2015 found that there were 347 protests and demonstrations in 2013, and a total of close to 1,200

11. Matsui, Hate Speech, supra note 10, at 443. 12. The term often used is "Korea residents," as a translation of the typically used Japanese terms I Ht A [South Korean in Japan] and iT H ) UA [North Korean in Japan], but I do not think the term "Korean residents" really captures the essence of who these people are and their status within Japan, as it does not distinguish between the third- and fourth-generation Koreans, most of whom cannot speak Korean and have never even been to the Peninsula, and those South Koreans who actually are just visiting or have recently immigrated to Japan. There is a huge literature on the Korean-Japanese community and the discrimination issues the community faces. See, e.g., CHANGSOO LEE & GEORGE DE Vos, KOREANS IN JAPAN: ETHNIC CONFLICT AND ACCOMMODATION (1981); Yasuaki Onuma, Interplay Between Human Rights Activities and Legal Standards of Human Rights: A Case Study on the Korean Minority in Japan, 25 CORNELL INT'L L.J. 515 (1992); tk BfKF, 'B H AU)S[ A4 [YASUAKI ONUMA, THE NATIONALITY AND HUMAN RIGHTS OF KOREANS IN JAPAN] (2004). 13. Japan'sFirst-Ever Hate Speech Probe Finds Rallies Fewer but Still a Problem, JAPAN TIMES, Mar. 30, 2016. See Jeremy Heard, Far-rightHate Speech againstKoreans (Osaka, Japan), YOUTUBE (July 10, 2013), https://www.youtube.com/watch?v=Gt6tRLgtBxY&feature-youtu.be (showing examples of such anti-Korea protests on YouTube). 462 HASTINGS CONSTITUTIONAL LAW QUARTERLY [Vol. 45:3 between April 2012 and September 2015.'4 This is likely a conservative estimate, and the number of instances of lower levels or more individual forms of hate speech is likely several multiples of this number. But these large and widely publicized protests attracted attention, particularly in Tokyo and Osaka, where there are large Korean-Japanese communities. The uptick in discriminatory and hostile attitudes towards Korean- Japanese in the early part of the decade was likely, in part, due to increased tensions with both North and South Korea, following the death of Kim Jong II of North Korea at the end of 2011, and the Liberal Democratic Party ("LDP") return to power in 2012. A combination of North Korean nuclear weapons ambitions, territorial disputes over uninhabited islands with South Korea, and the ongoing friction over how to resolve the Japanese wartime sex-slave issue (euphemistically referred to in Japan as the "Comfort Women" issue), increased tensions between Japan and the Koreas. This, in turn, inflamed attitudes towards the Korean-Japanese community. The first initiative to enact national legislation was actually taken by the Democratic Party of Japan during its short stint in government, but when the LDP returned to power it somewhat reluctantly responded with a draft bill of its own.15 The public profile of the issue was also further raised by a number of lawsuits against such anti-Korean groups as the Zaitokutai, for their conduct in protests against Korean-Japanese groups and institutions.1 6 But while several of these lawsuits were quite successful, obtaining damages in a handful of cases and even an injunction in a couple of high profile cases, 7 these cases also illustrated precisely why hate speech legislation is necessary. This is because in the absence of any hate speech law-that is, legislation that actually prohibits the expression of racist statements designed to foster hatred against minority groups such as the Korean-Japanese-other grounds for legal proceedings have to be found in either the criminal or civil law. And there are no such grounds that can be applied to limit or punish harmful

14. Japan'sFirst-Ever Hate Speech Probe Finds Rallies Fewer but Still a Problem, supra note 13. See also &WW, -4I '2L- A$r M, 3 } zi 28 4 [Ministry of Justice, Hate Report on the Inquiry into the Situation Regarding Hate Speech, Mar. 2016], http://www.moj.go jp/content/001201158.pdf (last visited Feb. 2018). 15. Mizuho Aoki, Opposition Slams LDP in PushingAntidiscrimination Bill, JAPAN TIMES, Aug. 6, 2015; Japan to Enact Bill to Counter Hate Speech, JAPAN TIMES, May 11, 2016. 16. Zaitokutai is the abbreviated name of the E W il f [the Association of Citizens Against the Special Rights of Koreans in Japan]. See Zaitokukai Ordered to Pay Damagesfor Defaming Korean Resident of Japan, JAPAN TIMEs, Sept. 27, 2016 (reporting on a decision by the Osaka District Court). The Kyoto District Court awarded damages against the Zaitokutai in 2013. 17. [Yokohama Dist. Ct., Kawasaki Branch], June 2, 2016, Hie 28(wo) No. 42; see also Tomohiro Osaki, Japanese Court Issues First-EverInjunction Against Hate-Speech Rally, JAPAN TIMES, June 3, 2016 (reporting on a decision by the Yokohama District Court). Spring 2018] STRIKING THE RIGHT BALANCE 463 statements made against identifiable groups in general, as opposed to attacks on specific individuals or institutions. This point is illustrated nicely in what are commonly referred to as the Kyoto North Korean School cases, which involved separate criminal and civil proceedings against the Zaitokutai. The cases related to a series of protests by the Zaitokutaiagainst a school in Kyoto for Korean-Japanese of North Korean heritage. The protests were particularly virulent, causing distress to young children in the school, and the demonstrations also caused damage to school facilities. At the request of the school the government prosecuted members of the Zaitokutai for forcible obstruction of business under Article 234 of the Criminal Code, damage to property under Article 261, and for criminal insult under Article 231. At the same time, the school commenced civil proceedings against members of the Zaitokutai for damages in tort, for defamation and insult, as provided for in the Civil Code. Both the criminal and the civil proceedings ultimately ended up at the Supreme Court, and the school prevailed in both, winning twelve million yen (approximately $100,000) in damages and a restraining order in the civil case.' The decision of the court of first instance in the civil proceeding garnered a great deal of attention because the court invoked and relied upon the CERD, an international human rights treaty that imposed obligations on Japan to provide protections from racial discrimination (I will return to review provisions of the CERD in more detail below).1 9 Specifically, the court held that the conduct and statements of the Zaitokutai came within the definition of discrimination in the CERD, which in turn provided the basis for finding liability for the tort of defamation.20 This reliance upon the CERD was seen as quite extraordinary by some commentators, and as a 2 reason for some optimism regarding the judicial approach to hate speech. 1 But the very same commentators also note that both the criminal and the civil

18. The criminal cases: [Kyoto Dist. Ct.], April 21, 2011, Hei 22(wa) no. 1257 and Hei 22 (wa) no. 1641; [Osaka High Ct.], Oct. 28, 2011, Hei 23 (u) no. 788; [Sup. Ct.], Feb. 23, 2012, Hei (a) no. 2009; and civil case: [Kyoto Dist. Ct.], Oct. 7, 2013, Hei 22 (wa) no. 2655, 2208 HANREI JIHO [HANJI] 74 (Japan); [Osaka High Ct], July 8, 2014, Hei 25 (ne) no. 3235; [Sup. Ct.], Dec. 9, 2014, Hei 26 (o) no. 1539, Hei 26 (ju) no. 1974. 19. CERD, supra note 5. 20. [Kyoto Dist. Ct.], Oct. 7, 2013, Hei 22 (wa) no. 2655, 2208 HANREI JIHO [HANJI] 74 (Japan). 21. For discussion of the case in English, see Junko Kotani, A Comment on Hate Speech Regulation in JapanAfter the Enactmentofthe Hate Speech EliminationAct of2016, 21 SHIZUOKA J.L. & POL. 1, 6-7 (2017); Matsui, Hate Speech, supra note 10, at 450-51; Ayako Hatano, Can Strategic Human Rights Litigation Complement Social Movements? A Case Study of the Anti-Hate Speech Movement in Japan, U. PA. ASIAN L. REV. (forthcoming 2018). 464 HASTINGS CONSTITUTIONAL LAW QUARTERLY [Vol. 45:3 grounds for this case are very much limited to statements and conduct directed toward and against specific individuals and institutions. These offenses and causes of action cannot be the basis for either prosecuting or suing anyone for statements intended to foster hatred generally against an identifiable group, no matter how hateful and racist the statements may be. 2 2 Indeed, the Kyoto District Court itself made clear in its judgment that there could be liability only when the hateful speech was directed towards specific persons or associations. 23 And so, in short, without hate speech legislation, the current criminal and civil codes can only provide redress for hate speech that can be characterized as defamation, insult, or threats against specific persons. This was the context in which the government finally and reluctantly enacted the current hate speech legislation, which, as we will see below, neither prohibits hate speech nor punishes it. But before examining the legislation itself, it is perhaps worth noting how the law implicates a more general debate about Japanese law-worth noting because certain perspectives from this debate may indeed inform the critique of the new hate speech legislation. The debate relates to the manner in which Japanese laws, even criminal laws, often have either very minimal sanctions, or often no punishment at all. Some Western scholars have argued that Japanese law is successful precisely because it relies on methods other than formal sanction in order to mobilize compliance.24 John Haley is perhaps most identified with this theory. He famously argued that one of the functions of this weak form of law was to itself create and shape positive social norms. In his view, Japanese criminal law was a significant factor in explaining Japan's relatively low crime rate and low recidivism rate, precisely because of this relatively weak formal enforcement and sanctions regime, and reliance upon informal societal norms and institutions to mobilize compliance.25 Moreover, this was evidence for an observation that Japanese law more generally, as compared to Western legal systems, often mobilizes compliance through the influencing of informal societal norms rather than through the formal imposition of sanctions.26 Many Japanese scholars, such as Setsuo Miyazawa, have argued that these explanations tend to miss some of the negative aspects and

22. Kotani, supra note 21, at 7; Matsui, Hate Speech, supra note 10, at 452. 23. [Kyoto Dist. Ct.], October 7, 2013, Hei 22 (wa) no. 2655, 2208 HANREI JIHO [HANJI] 74 (Japan); Kotani, supra note 21, at 7. 24. JOHN OWEN HALEY, AUTHORITY WITHOUT POWER: LAW AND THE JAPANESE PARADOX (1991); JOHN OWEN HALEY, THE SPIRIT OF JAPANESE LAW (1998). 25. HALEY, AUTHORITY WITHOUT POWER, supra note 24, at 138; HALEY, SPIRIT OF JAPANESE LAW, supra note 24, at 88-89. 26. HALEY, AUTHORITY WITHOUT POWER, supra note 24, at 138-140. Spring 2018] STRIKING THE RIGHT BALANCE 465 ramifications of such leniency. They do not dispute the descriptive accuracy of Haley's observation that Japanese law is often comparatively lenient, but they challenge the normative implications. Miyazawa, for instance, has argued that these efforts to explain some of the apparently successful aspects of Japanese law as being the result of the law's weak sanctions and enforcement regime, helped to obscure other serious problems within the legal system. In short, they argued that the lack of enforcement mechanisms and sanctions in Japanese laws can often disguise and sometimes even contribute to a disregard for, and even a direct violation of, fundamental individual rights. 27 This of course plays out in different ways in different areas of Japanese law. In the area of human rights, however, the apparent leniency and lack of enforcement of what human rights law exists, operates to negatively affect the very people the law should be protecting. As will be discussed below, there has been a persistent failure to develop legislation that not only defines and prohibits the violation of fundamental rights (such as the rights of minorities not to be discriminated against), but when laws are contemplated or enacted, they generally lack any meaningful sanctions and remedies in the event of a breach. As such this leniency towards perpetrators of harm has contributed to the continued systemic violation of the rights of minorities, the people the law is supposed to protect. 28 The recent hate speech laws reflect this paradigm. The legislation was passed to ostensibly protect minorities, particularly members of the large Korean-Japanese community in Japan, in the face of rising levels of virulent hate speech that will likely contribute to increased discrimination, and possibly even persecution and violence. Yet the laws lack any form of sanction or mechanisms of enforcement. They thus leave the vulnerable group exposed and unprotected, notwithstanding that the very enactment of the law acknowledges their plight. What is more, the passage of the law may create a false sense that action is being taken to protect the vulnerable, thus helping mask the continued violations, and provide an excuse for taking no further action. Indeed, Frank Upham has argued that it has been a typical stratagem of Japanese governments in the past to pass toothless legislation in response to civic unrest and social rights litigation, as a means of defusing the situation and dividing the opposition.29 In this sense, the hate speech laws tend to provide

27. Others have argued that Haley's analysis overstates the extent to which Japanese law is comparatively lenient or bereft of power. See, e.g., David T. Johnson, Review Essay, Authority with Power: Haley on Japan'sLaw and Politics, 27 LAW & Soc'Y REv. 619 (1993). 28. See infra notes 47-66 and associated text. 29. UPHAM, supra note 10; and comments of Frank Upham at the American Association of Law Schools Annual Conference, San Diego, Jan. 4, 2018. 466 HASTINGS CONSTITUTIONAL LAW QUARTERLY [Vol. 45:3 yet another example of the problem that Miyazawa addressed, and illustrates how lenient laws in Japan can at times facilitate the violation of rights.

B. The Hate Speech Legislation The government finally enacted a law on hate speech in May 2016.30 During the debate of the bill in the Diet it was roundly criticized as being excessively narrow as well as toothless, and yet it was further weakened during the process of debate and revision. In the end, the law comprised of only seven short articles, and it is indeed deserving of some hard questions. To begin, it notes in the preamble that there has been a rise in discriminatory speech and behavior in recent years, and "declares that such unfair discriminatory speech and behavior will not be tolerated," and that the law was enacted to "spread awareness among the general public . .. to strengthen efforts to eliminate unfair discriminatory speech and behavior." 31 This is all very general and hortatory, and the remaining articles, under the heading "Basic Measures" (which are presumably designed to help achieve these vague aspirations), are similarly vapid. Article Five contemplates the national and local governments establishing some form of unspecified consultation system for victims of hate speech, which is to help "prevent and resolve disputes in this regard."32 Article Six requires the national and local governments to implement educational activities to help eliminate hate speech. Article Seven requires both national and local governments to "spread awareness among the general public about the need to eliminate" hate speech, and to "implement public relations activities for the purpose of furthering understanding" of this need.33 That is it. There is no actual prohibition of either the creation or the dissemination of hate speech, nor any sanctions whatsoever for the communication of hate speech. Indeed, the law never addresses the perpetrators of hate speech directly at all-all of the provisions are directed to imposing obligations on the national and local governments to raise awareness and understanding within society, and on the general public at large to improve its understanding of the issues. The law

f, http://www.moj.go.jp/content/001184402.pdf (last visited Feb. 2018). The law came into force on June 3, 2016, as Law No. 68, 2016; an English translation, The Act on the Promotion of Efforts to Eliminate Unfair Discriminatory Speech and Behavior against Persons Originating from Outside Japan (provisional translation), http://www.moj.go jp/ENGLISHjmjinkenO4_00001.html (last visited Feb. 2017) [hereinafter Discriminatory Speech Law]. 31. Discriminatory Speech Law, supra note 30, at preamble. 32. Id. at art. 5. 33. Id. at art. 6-7. Spring 2018] STRIKING THE RIGHT BALANCE 467 is not directed at, nor does it even refer to, the individuals or entities engaging in the communication of hate speech. It is for this reason that there has been such criticism of the law as being ineffectual.34 The other cause for criticism of the new law is that it defines hate speech in a manner that is both vague and yet excessively narrow. Article One of the law defines the subject of the law as "unfair discriminatory speech and behavior against persons originating from outside of Japan."3 5 It goes on to define each of the elements of this clause. Unfair discriminatory speech and behavior is defined as "discriminatory speech and behavior to incite the exclusion... from the local community ... [by] openly announcing to the effect of harming the life, body, freedom, reputation or property of, or to significantly insult" such persons from outside of Japan.36 The addition of "significantly insult" was only added at the last minute, and without it the definition could have been construed even more narrowly to only that speech that threatens to harm the life, person, liberty, property or reputation of the targets of such speech.37 But even as it stands now, it is not clear that the definition is really trying to prevent the kind of speech that can foster hatred and discrimination against minorities. Indeed, it is not entirely clear precisely what kind of speech is contemplated. As will be discussed below, the literature on hate speech and the jurisprudence in other countries recognizes that there is significant harm to minorities caused by speech that vilifies, denigrates, demeans, and even dehumanizes identifiable groups within society, and thereby fosters hatred, prejudice, and increased discrimination against visible minorities within a society. Such speech need not (though it may) rise to the level of threatening their life, body, freedom, or property in order to cause this significant harm.38

34. Mizuho Aoki, Opposition Slams LDP in PushingAntidiscrimination Bill, JAPAN TIMES, Aug. 6, 2015, https://www japantimes.co.jp/news/2015/08/06/national/politics-diplomacy/opp osition-slams-ldp-pushing-antidiscrimination-bill/; Japan to Enact Bill to Counter Hate Speech, JAPAN TIMES, May 11, 2016; see generally t Fi$&, F-4 b J:3A0 n 0 v v [SHINJI HIGAKI, A CONSTITUTIONAL CONSIDERATION OF HATE SPEECH REGULATION: THE FREEDOM OF EXPRESSION DILEMMA] (2017); ri) I IE A, F*AO. n LA0 7 , Ati$M 2015t2-9 (360 T) [Masato Ichikawa, Freedom ofExpression andHate Speech, 2:360 RITSUMEIKANHOGAKU 122 (2015)]; Matsui, Hate Speech, supra note 10. 35. Discriminatory Speech Law, supra note 30, at art. 1 (Japan). 36. Discriminatory Speech Law, supra note 30, at art. 1 (Japan). 37. See infra Part II-B for discussion of the "fighting words" and "incitement to violence" categories in American freedom of speech doctrine. 38. See, e.g., RICHARD DELGADO & JEAN STEFANCIC, MUST WE DEFEND NAZIS?: HATE SPEECH, PORNOGRAPHY, AND THE NEW FIRST AMENDMENT (1997); JEREMY WALDRON, THE HARM IN HATE SPEECH (2012); HIGAKI, supra note 34; Richard Delgado, Words that Wound, 17 4689 HASTINGS CONSTITUTIONAL LAW QUARTERLY [Vol. 45:3 The second element of the definition, the term "persons from outside of Japan," was defined as "persons originating exclusively from a country or region other than Japan or their descendants and who are lawfully residing in Japan." 39 The first major criticism is that the law thereby only purports to be concerned with protecting foreigners, or their descendants. The focus on foreigners and their descendants likely reflects the fact that much of the controversy over hate speech in recent years has been driven by discrimination against members of the Korean-Japanese community. In one sense the law, while ostensibly for the protection of members of the Korean- Japanese community, with its exclusive focus on "foreigners" merely serves to further emphasize the alienation and foreignness of a people who have been in Japan for generations. But it also clearly excludes and ignores some other important identifiable groups within Japan, most obviously members of the Burakumin and Ainu communities, who have for generations been the victims of systemic discrimination, including hate speech.40 What is more, by further limiting the scope of the law to only those foreigners who are "lawfully in Japan," the law tends to signal to society that hateful communication is acceptable if directed at refugee claimants or other foreigners who have overstayed their visa term, or whose immigration status is otherwise not valid. This would include a large percentage of the thousands of children of uncertain parentage who have been born and raised in Japan, but have been denied Japanese nationality, and many of whom are stateless. 4' The Japan Lawyers Network for Refugees, among other groups, were critical of the new law for this very reason.42 Thus, in short, by definition the law only purports to address speech directed against a narrow

HARV. C.R.-C.L. L. REV. 133 (1982); Mari J. Matsuda, Public Response to Racist Speech: Consideringthe Victim's Story, 87 MICH. L. REv. 2320 (1989); Mayo Moran, Talking About Hate Speech: A RhetoricalAnalysis ofAmerican and CanadianApproaches to the Regulation ofHate Speech, 1994 Wisc. L. REv. 1425 (1994); for foreign jurisprudence, see, e.g., R. v. Keegstra [1990] 3 S.C.R. 697 (Supreme Court of Canada); for research on the harms and effects of hate speech and incitement at the international level, see RICHARD ASHBY WILSON, INCITEMENT ON TRIAL: PROSECUTING INTERNATIONAL SPEECH CRIMES (2017). 39. Discriminatory Speech Law, supra note 30, at art. 1. 40. For discussion of the Burakumin issues, see, e.g., Japan Federation of Bar Associations, "Japan Federation of Bar Associations Report on Response to the Seventh, Eighth, and Ninth Report of the Japanese Government of the International Convention on Elimination of All Forms of Racial Discrimination," Mar. 19, 2014, at 72-83 [hereinafter JFBA Report]. 41. See, e.g., Chen Tien-shi, Statelessness in Japan: Management and Challenges, 21 J. POPULATION AND SOC. STUD. 70 (2012). 42. Tomohiro Osaki, Diet Debates Hate-Speech Bill That Activists Call Narrow and Toothless, JAPAN TiMES, Apr. 19, 2016, https://www.japantimes.co.jp/news/2016/04/19/national/politics- diplomacy/diet-debates-hate-speech-bill-activists-call-narrow-toothless/#.WlkiMKhKvD4. Spring 2018] STRIKING THE RIGHT BALANCE 469 subset of vulnerable identifiable minorities within Japan; and primarily addresses speech that threatens harm against the life, person, liberty, property, or reputation of such persons. And in substance, the law does not actually prohibit or prescribe any punishment for such speech. In addition to this legislation, there have been moves to enact municipal or prefectural ordinances aimed at limiting hate speech. Indeed, some might argue that the enactment of the national legislation helped provide the impetus for these developments at the regional level. But some of the local efforts predated the national government initiative, and so the influence may well have run in the other direction. Or they both may have been a response to increased pressure to deal with a problem that was gaining more prominence. Osaka, for instance, promulgated an ordinance that came into effect in July 2016, which both defined hate speech more broadly and provided for greater enforcement mechanisms than the government legislation.43 It defined hate speech as any communication that defames and aims to exclude a particular group based on race or ethnicity, and disseminates such information to large numbers of people through such media as the internet. It included a complaints process, with a panel established to consider complaints, and vested with the power to take such action as publishing the names of those found to have engaged in hate speech, and request internet servers to remove offending material from client websites. 44 But here too, the ordinance lacks the typical form of legal sanctions available to local governments, such as the levying of fine. The lack of sanctions notwithstanding, the naming and shaming contemplated by the Osaka ordinance would, it has been suggested, provide the basis for municipal authorities to subsequently deny offending groups the licenses and approvals necessary to hold demonstrations and rallies in public areas, and make it difficult for such groups to rent other facilities from private companies. 45 In this way it might provide some check against the conduct of anti-Korean groups.

43. )OVrI 4 1 7 i [Osaka City Ordinance on Treatment of Hate Speech], Osaka City Ordinance No. 1, Jan. 18, 2016 (Japan), unofficial translation available in Koji Higashikawa, Japan 's Hate Speech Laws: Translations of the Osaka City Ordinance and the National Act to Curb Hate Speech in Japan, 19 ASIAN-PAC. L. & POL'Y J. (2017), and an unofficial translation by the Asia-Pacific Human Rights Information Center is also available online at https://www.hurights.or jp/archives/racism-elimination/osaka city hate%20speech _ordinance english.pdf; see also, Eric Johnston, Osaka Enforces Japan 's First Ordinance Against Hate Speech, Threatens to Name Names, JAPAN TIMEs, July 1, 2016, https://www.japan times.co.jp/ news/2016/07/0 1/national/crime-legal/osaka-enforces-japans-first-ordinance-hate-speech-threaten s-name-names/#.WlkioahKvD4. 44. Osaka City Ordinance on Treatment of Hate Speech, supra note 43. 45. Johnston, supra note 43. 470 HASTINGS CONSTITUTIONAL LAW QUARTERLY [Vol. 45:3 It is still too early to tell, given that the new law was enacted little more than a year ago, but some might be inclined to think that it may have a positive impact. It has been argued, for instance, that the Osaka court decision ordering an injunction against an anti-Korean protest in the summer of 2016 may have been influenced by the new law and indicates that the new hate speech law will have a positive influence on the future judicial treatment of hate speech. That is, that the law, and particularly its definition of "discriminatory speech and behavior," could help shape the manner in which courts interpret and apply various other relevant laws that may provide a civil cause of action. Thus, according to this argument, while the hate speech law itself provides for no sanctions or penalties against the perpetrators of hate speech, or any remedies for the victims of hate speech, it may nonetheless bolster and make possible lawsuits grounded in tort and other legal regimes.46 This, of course, plays into Haley's famous and favorable interpretation of how Japanese laws exercise authority even while declining to exert or enforce the power of the state. But I would suggest that this is altogether too optimistic a view when the legislation is placed into a larger context. The entire effort needs to be considered within a broader understanding of both equality rights and freedom of expression in Japan.

C. Equality Rights and Discrimination Hate speech is only one small slice of the kind of discrimination that minorities in Japan face. And in contrast to the narrowly defined category of persons protected by the hate speech law, those who face systematic discrimination in Japan include: foreigners generally, as well as Korean- Japanese, Taiwanese-Japanese (another significant community of people who, like Korean-Japanese, are second, third- or fourth-generation residents of Japan); the Burakumin; the Ainu; the Ryukyu; racial and ethnic minorities; persons of Japanese descent who have immigrated from Brazil and Peru; children born to persons out of wedlock; the disabled; persons with such diseases and health conditions as HIV/AIDS, Hanson's Disease, and even epilepsy; and women, to name only the most prominent. The United Nations Human Rights Committee, the institutional body that oversees the implementation and enforcement of the ICCPR, has for decades criticized Japan for its failure to develop laws and institutions to prohibit discrimination and to provide a legal framework enabling victims to seek redress. As recently as 2014, the Human Rights Committee, in its

46. Kotani, Comment on Hate Speech, supra note 21, at 10. 47. ICCPR, supra note 4. Spring 2018] STRIKING THE RIGHT BALANCE 471 "Concluding Observations on the Sixth Periodic Report of Japan," reiterated its concerns. 8 It pointedly noted that Japan had not yet acted to address many of the concerns expressed by the Committee in the past. In particular, it noted that the rights provided for in the ICCPR, including the right to equality and not to be discriminated against, were not generally protected or justiciable in Japanese courts, and that Japan had still not developed any national human rights legislation or institutions. Indeed, the report noted with particular regret that the Japanese government had abandoned an attempt in 2012 to pass a bill establishing a new Human Rights Commission, and that no further action had been attempted since. 49 The report noted the continued inequality and discrimination faced by women, foreigners, gays, and lesbians, and rising levels of hate speech and discrimination directed against minority groups such as Koreans, Chinese, Burakumin, and indigenous peoples such as the Ainu and Ryukyu.o Similarly, the Committee on the Elimination of Racial Discrimination, which is the institutional body that oversees the implementation and enforcement of the CERD, has been consistently critical of Japan's failure to respect and enforce the rights in the convention, since Japan acceded to the Convention in 1995.51 The CERD imposes very specific obligations on states in relation to hate speech, providing that:

Article 4. - States Parties condemn all propaganda and all organizations which are based on ideas or theories of superiority of one race or group of persons of one colour or ethnic origin, or which attempt to justify or promote racial hatred and discrimination in any form, and undertake to adopt immediate and positive measures designed to eradicate all incitement to, or acts of, such discrimination and, to this end, with due regard to the principles embodied in the Universal Declaration of Human Rights and the rights expressly set forth in article 5 of this Convention, inter alia: (a) Shall declare an offence punishable by law all dissemination of ideas based on racial superiority or hatred, incitement to racial discrimination, as well as all acts of violence or incitement to such acts against any race or group of persons of another colour or ethnic origin, and also the provision of any assistance to racist activities, including the financing thereof;

48. Human Rights Comm., Concluding Observations on the Sixth Periodic Report of Japan, U.N. Doc. CCPR/C/JPN/CO/6 (2014) [hereinafter HRC Report, 2014]. 49. Id. at paras. 6-7. 50. Id. at paras. 8-12, 26-29. 51. CERD, supra note 5. 472 HASTINGS CONSTITUTIONAL LAW QUARTERLY [Vol. 45:3 (b) Shall declare illegal and prohibit organizations, and also organized and all other propaganda activities, which promote and incite racial discrimination, and shall recognize participation in such organizations or activities as an offence punishable by law; (c) Shall not permit public authorities or public institutions, national or local, to promote or incite racial discrimination.52

Japan registered a reservation to this article of the convention at the time it acceded to it, limiting the scope of the obligation to the extent Japan found it consistent with the constitutional right to freedom of expression. The CERD Committee, however, has criticized Japan for failing to develop specific and comprehensive laws prohibiting both direct and indirect racial discrimination, in accordance with its obligations under Article Four of the CERD, as well as its failure to establish national human rights institutions in accordance with the Paris Principles. 5 3 What is more, the CERD Committee has consistently and repeatedly criticized Japan for maintaining its reservation, arguing that the obligations under the CERD can be reconciled with robust rights to freedom of expression.54 As should be quite clear from the description of the new hate speech law above, it does not fulfill the obligations Japan would have under the CERD but for the reservation. It does not constitute a law that "declares it an offense punishable by law all dissemination of ideas based on racial superiority or hatred, incitement to racial discrimination .. .," nor does it "declare illegal and prohibit organizations, and also organized and all other propaganda activities, which promote and incite racial discrimination." It is worth noting that it does not even appear to conform to the CERD definition of racial discrimination in its own definition of hate speech. The CERD defines "racial discrimination" quite broadly, as including "any distinction, exclusion, restriction or preference based on race, colour, descent, or national or ethnic origin which has the purpose or effect of nullifying or impairing the recognition, enjoyment or exercise, on an equal footing, of human rights

52. CERD, supra note 5, at art. 4. 53. Comm. on the Elimination of Racial Discrimination, Concluding Observations on the Combined Seventh to Ninth Periodic Reports of Japan, U.N. Doc CERD/c/jpn/co/7-9, paras. 8-9 (2014) [hereinafter CERD Report 2014]; see also JFBA Report, supra note 40. The Paris Principles were established at the first International Workshop on National Institutions for the Promotion and Protection of Human Rights in Paris, Oct. 7-9, 1991, and adopted by the U.N. General Assembly in Resolution 48/134 of 1993. 54. CERD Report 2014, supra note 53, at para. 10. Spring 2018] STRIKING THE RIGHT BALANCE 473 and fundamental freedoms. ... "" The limitation of the new Japanese hate speech law to expression directed at foreigners and their decedents lawfully in Japan, does not begin to address the obligation in the CERD. The CERD Committee, in its recent report on Japan, also reviewed the continuing discrimination against foreigners, women, ethnic and racial minorities, the Burakumin, the Ainu, Muslims, and other religious minorities, among others. The Committee also, like the Human Rights Committee, focused on the particular problem of hate speech and hate crimes. It noted the increase in the incidence of hate speech and discriminatory behavior, and criticized Japan for its failure to implement laws designed to protect minorities from hate speech, to investigate and prosecute those who engage in such conduct, and to punish public officials and politicians who disseminate hate speech.56 And for those who may be inclined to dismiss the criticisms of the institutions of international human rights law as being not entirely relevant to a debate over the domestic laws of Japan, it should be noted that the Constitution of Japan, in the chapter titled "Supreme Law," provides that treaties to which Japan is a party, in addition to customary international law, "shall be faithfully observed."57 This provision has been interpreted, by the government, the courts, and legal scholars, as meaning that treaties to which Japan is a party are directly and automatically incorporated as the law of the land, without requiring any implementing legislation, and moreover such law prevails over any conflicting statutes.5 " This was indeed the basis upon which the Kyoto District Court referred directly to the CERD in the Kyoto North Korean School cases discussed above. The repeated observations of the U.N. committees are just part of a larger body of evidence that reflects the fairly well established fact that the rights of minorities in Japan are not well protected by the legal system. 59 Indeed, as I have argued elsewhere, even the constitutional right to equality and not to be discriminated against, provided for in Article 14 of the

55. CERD, supra note 5, art. 1. It should be noted that art. 1(2) does provide that the definition does not apply to distinctions, exclusions, restrictions, or preferences made by the state party between citizens and noncitizens-so denying noncitizens the right to vote, for instance, does not violate the convention. 56. CERD Report, supra note 53 , at paras. 10-11. 57. Constitution of Japan, 1947, art. 98, para. 2 (Japan). 58. See, e.g., YUJI IWASAWA, INTERNATIONAL LAW HUMAN RIGHTS AND JAPANESE LAW (1998), at 27, and chap. 3. The Japanese Supreme Court has also directly interpreted treaties. See, e.g., 28Minshu 1331 (Sept. 26, 1974); 31Minshu 511 (June 28, 1977). Lowercourts have similarly directly implemented provisions of the ICCPR. See IWASAWA, at 51c56. 59. See, e.g., JFBA report, supra note 40. 474 HASTINGS CONSTITUTIONAL LAW QUARTERLY [Vol. 45:3 Constitution of Japan, is not rigorously enforced by the judiciary.6 0 What is more, there has been considerable reluctance on the part of successive governments to establish legal protections against discrimination. During the Socialist Party's brief turn as part of a coalition government in the 1990s, a Law on the Promotion of Measures for Human Rights Protection was developed, and finally enacted in 1996-the year after Japan acceded to the 6 CERD. 1 It too lacked any prohibitions, sanctions, or enforcement mechanisms, but it was seen as the first step towards developing a human rights regime. But it was repealed in 2002, after the LDP returned to power. In 2002 and again in 2005 the Diet considered but failed to enact new human rights legislation. 62 Similarly, a Human Rights Commission Bill was developed in 2012, but it was allowed to die when the Diet was dissolved in November, and never re-submitted for consideration.6 3 This reluctance is not unique to politicians at the national level. Local governments, too, have been slow to enact ordinances to protect the equality rights of minorities. What is more, those that have taken action on the issue have often come under intense pressure to reverse course. In 2006, for instance, Tottori Prefecture enacted an antidiscrimination human rights ordinance, the Ordinance Regarding the Promotion and Procedure for the Restitution for Human Rights Violations, only to be forced by conservative groups and media outlets to revoke it.64 As mentioned above, some local governments, such as that of Osaka, have more recently enacted hate speech ordinances; however, there continues to be considerable skepticism and suspicion regarding human rights legislation designed to protect minorities from discrimination, particularly when the measures are viewed as primarily benefiting Korean-Japanese individuals. At best, this tends to indicate government indifference towards the rights of minorities more generally. Thus, when examining the hate speech law issue within that context, there is good reason to believe that the government's approach simply does not take seriously the rights of victims

60. Martin, Glimmers ofHope, supra note 10. 61. Law No. 120, 1996. 62. A [HumanRights ProtectionBill], House of Representatives, 154th Session, 2002, http://www.moj.go jp/content/000104841.pdf. For discussion of this, see also Matsui, Hate Speech, supra note 10, at 436-37. 63. A V W 4 [Human Rights Commission Bill], http://www.moj.go jp/content/ 000104039.pdf. Reference is made to its demise in the CERD Report, supra note 53, para. 7, and the HRC Report, supra note 48, at para. 9. 64. Debito Arudou, How to Kill a Bill: Tottori's Human Rights Ordinance is a Case in Alarmism, JAPAN TIMES, May 2, 2006, https://www japantimes.co jp/community/2006/05/02/ issues/how-to-kill-a-bill/#.WmUyeZM-fOQ. Spring 2018] STRIKING THE RIGHT BALANCE 475 of hateful expression, or the nature of harm that such speech is likely to cause. There is little acknowledgement that such expression is likely to foster prejudice and incite discrimination.65 Moreover, there is little apparent understanding that such speech may possibly cause even more harmful behavior towards minorities, as well as have a broader negative impact on the values of equality and tolerance in society.6 6 Now, viewed through the lens of Haley's "authority without power" paradigm, one might argue that the hate speech law will nonetheless help to shape cultural and social norms by its mere enactment, and through the awareness raising and educational programs that the law requires the national and local governments to implement. Moreover, as mentioned earlier, it may influence the courts in how they interpret and apply civil laws giving rise to other causes of action, such as defamation, in law suits against the perpetrators of hate speech. Over time these influences of the law will operate to reduce the incidence of hate speech. Perhaps there is some merit to such claims, but the jury will be out for some time before we know if that is right. In the meantime, however, people will continue to suffer harm. And the question has to be asked: why shy away from enacting real prohibitions on a narrow band of hate speech, and establishing real sanctions for those who engage in hateful communication that is likely to foster discrimination and otherwise cause harm to identifiable minorities? The apparent leniency towards the perpetrators of hate speech is at the expense of the victims, whose rights are being discounted or even ignored. Japan has obligations under international law, and arguably under the Constitution, to protect those rights, and yet it has consistently resisted doing so. I say arguably because of course Article 14 of the Constitution first and foremost governs the relationship between the government and the individual, not between private entities and the individual. But having said that, the courts have on occasion interpreted labor law and other statutes in a manner that explicitly invokes the requirement to conform with the values enshrined in Article 14.67 It is

65. Indeed, the government, in its periodic report to the CERD Committee, formally took the position that it "does not believe that, in present-day Japan, racist thoughts are disseminated and racial discrimination is incited, to the extent that the withdrawal of its reservations or legislation to impose punishment against dissemination of racist thoughts and other acts should be considered even at the risk of unduly stifling legitimate speech." Comm. on the Elimination of Racial Discrimination, Seventh to Ninth Combined Periodic Reports by the Government of Japan Under Article 9, U.N. Doc CERD/C/JPN/7-9, para. 84 (2013). 66. For more on the harms caused by hate speech, see supra note 38. 67. See, e.g., Supreme Court Judgment (Mar. 24, 1981), 35 Minshii 2 300 (Nissan Motors case) (holding that "a lower compulsory retirement age for women than for men constitutes discrimination against women based solely on their gender and is irrational discrimination invalid 476 HASTINGS CONSTITUTIONAL LAW QUARTERLY [Vol. 45:3 arguable, therefore, that the government has an obligation under Article 14 not only to treat individual equally, but also to enact laws to ensure that "all of the people" enjoy a right to equality and are not discriminated against in their interactions with private entities and persons. Considering the issue in the broader context of the failure to protect minorities from discrimination more generally, it is difficult not to incline towards rather unfortunate and negative conclusions regarding the likely explanations for this reluctance. 68 But the more positive explanation that is typically offered by defenders of the government approach, is that it is for reasons of protecting the right to freedom of expression. In other words, the approach is actually motivated by an impulse to defend rights, not to ignore or neglect them. And to be fair, the tension between hate speech and freedom of speech has been a challenge for many constitutional systems, and so we need to explore more closely the argument in the Japanese context.

D. Freedom of Expression The strongest objection to a more rigorous and enforceable hate speech law is that it can limit or violate the right to freedom of expression. This is the objection that has been made by both politicians and public figures in debate over the legislation, and indeed by scholars writing on the issue. 69 The Constitution of Japan includes a robust right to freedom of expression in Article 21, which provides that: "Freedom of assembly and association as well as speech, press and all other forms of expression are guaranteed. No censorship shall be maintained, nor shall the secrecy of any means of communication be violated."" A law that prohibited certain forms of expression, based on the content of that expression, would on its face appear to limit or infringe this right. And, as mentioned earlier, when Japan acceded to the CERD in 1995, it registered a reservation to the effect that:

In applying the provisions of paragraphs (a) and (b) of article 4 of the [said Convention] Japan fulfills the obligations under those provisions to the extent that fulfillment of the obligations is compatible with the

under Article 90 of the Civil Code . . . Article 1-2"). For an English translation of this case, under the title Nissan Motors, Inc. v. Nakamoto, see LAWRENCE W. BEER AND HIROSHI ITOH, THE CONSTITUTIONAL CASE LAW OF JAPAN, 1970 THROUGH 1990 (1996) at 179-8 1. 68. Others share this inclination. See, e.g., Matsui, Hate Speech, supra note 10, at 433-36. 69. For review of the debate, see, e.g., Ichikawa, supra note 34 and HIGAKI, supra note 34. 70. CONST. OF JAPAN, art. 21. Spring 2018] STRIKING THE RIGHT BALANCE 477 guarantee of the rights to freedom of assembly, association and expression and other rights under the Constitution of Japan .... .1

Article 4(a) and (b) of the CERD, as explained above, imposes explicit obligations on state parties to implement laws that prohibit and punish both expression that promotes and incites racial discrimination, as well as organizations and propaganda activities that promote or incite racial discrimination. So, the key question is whether such laws would be compatible with the rights to freedom of expression and association in Article 21 of the Constitution of Japan. And, as discussed in the Introduction, this is not only a question for Japan-finding the right balance between respecting and enforcing the rights to freedom of expression and association on the one hand, while on the other hand protecting the right to be treated equally under the law and not to be discriminated against, is the challenge facing many constitutional democracies in trying to craft laws that govern hate speech. In order to address this question in the context of Japan, however, it is necessary to provide some explanation of the right to freedom of expression in Japan. First, freedom of expression arguments against hate speech laws in Japan, particularly those made by politicians from the governing LDP, are difficult to take entirely at face value. More precisely, it is difficult to interpret such arguments as reflecting good faith concerns for the protection of the constitutional right to freedom of expression. Rather, they are more likely convenient pretexts to help justify the government's refusal to implement more rigorous laws to defend the rights of minorities. Such skepticism arises from the fact that, while in power, the LDP has shown scant regard for the right to freedom of expression in its enactment of other laws. Most recently, in the same period in which the hate speech law was being debated, it passed laws and engaged in conduct that was widely condemned for undermining the right to freedom of expression. The most significant instance of this was the passage of a state secrets law, the Act on the Protection of Specially Designated Secrets, 72 which criminalized and established severe prison sentences for the disclosure of a vastly expanded

71. CERD, supra note 5, reservationby Japan (Dec. 15, 1995). 72. i 'T-)25 12A 3 H $f 108, [The Law Relating to the Protection of Specially Designated Secrets, Law No. 108, 2013] http://law.e- gov.gojp/htmldataIH25IH25HO108.htm [hereinafter Special Secrets Law] (English translation at http://wwwjapaneselawtranslation.go jp/law/detail/?id=2543&vm=04&re=01). 4789 HASTINGS CONSTITUTIONAL LAW QUARTERLY [Vol. 45:3 range of information vaguely defined as "national secrets."7 3 The law was severely criticized within Japan, not only by various organizations of journalists, but by the Japanese Federation of Bar Associations and other advocacy and human rights groups. There was also criticism from outside Japan. The U.N. Human Rights Committee, in its Concluding Observations on the periodic report of Japan under the ICCPR, singled out the Specially Designated Secrets Law as likely violating the rights to freedom of expression and freedom of the press provided for in Article 19 of the ICCPR. It noted that, in particular, the law contained vague and overly broad designations of the kind of information that would be subject to the law, coupled with severe criminal sanctions, which together would combine to create a significant chilling effect on the activities of journalists and human rights advocates. While the enactment of the Specially Designated Secrets Law stands out as the most obvious example, the government has engaged in many other activities over the last five years that have been criticized for interfering with and suppressing both freedom of the press and freedom of expression- frequently to influence reporting on such issues as the government efforts to reinterpret Article Nine of the Constitution, and the official failures regarding the Fukushima nuclear disaster.7 6 Concerns have been expressed in many quarters that press freedom in Japan has declined in the last five years as a result of these developments.

73. For analysis of the law, see, e.g., Lawrence Repeta, A New State Secrecy Law for Japan?, 11 ASIA-PACIFIC J.: JAPAN Focus, Oct. 18, 2013, http://apjjf org/2013/11/42/Lawrence-Repeta/40 11/article.html; Lawrence Repeta, Japan's 2013 State Secrecy Act-the Abe Administration's Threat to News Reporting, 12 ASIA-PACIFIC J.: JAPAN FOCUS, Mar. 3, 2014, http://apjjf org/2014/ 12/10/Lawrence-Repeta/4086/article.html. 74. See Repeta, Japan's2013 State Secrecy Act, supra note 73, for a compilation of sources, but for the JFBA analysis see, e.g., JFBA, Statement Opposing the Special Secrets Bill and Calling for a Full-Scale Review on the System for a Protection of Secrecy Law in Accordance with the Tshwane Principles, Nov. 15, 2013, http://www.nichibenren.or jp/en/document/statements/year/ 2013/131115.htil. 75. HRC Report, 2014, supra note 48, at para. 23. 76. See, e.g., Anchors Away: Media Freedom in Japan, ECONOMIST, Feb. 18, 2016, http://www.economist.com/news/asia/21693269-criticism-government-being-airbrushed-out- news-shows-anchors-away; Is Japan'sPublic Broadcaster Under Threat, BBC NEWS, Mar. 20, 2014, http://www.bbc.com/news/world-asia-26403639; News Giant in Japan Seen as Being Compromised, N.Y. TIMES, Feb. 2, 2014, https://www.nytimes.com/2014/02/03/world/asia/news- giant-in-japan-seen-as-being-compromised.html. 77. See, e.g., Reporters Without Borders, Japanese Government Threatens to Tighten Grip on Broadcasters, Feb. 25, 2016, http://www.ifj.org/nc/news-single-view/backpid/1/article/japan ese-governnents-threatens-to-tighten-grip-on-broadcasters/; Reporters Without Borders, RSF Spring 2018] STRIKING THE RIGHT BALANCE 479 What is more, it cannot be said that the government's reluctance to enact a more restrictive and effective hate speech law was driven by some well-founded concern that the law might be struck down by the courts. The passage of the broad, restrictive, and highly punitive Specially Designated Specially DesignatedSecrets Law alone stands against that proposition. But more significantly, freedom of expression has never been rigorously enforced by the courts. Indeed, in the seventy years since the Constitution was promulgated, the Supreme Court has never struck down a law or found a government policy or action unconstitutional by reason of violating Article 21 of the Constitution. And that is not because there have been no laws challenged as violating Article 21, or that all laws in Japan have been uniformly enacted so as to respect and comply with the rights in question. In fact, there has been no shortage of cases challenging laws as violating freedom of expression and association. In many of those cases, the laws in question-laws enacted by LDP governments-were highly suspect in terms of their limits on freedom of speech. As Shinegori Matsui (among many others) has argued, the Supreme Court has simply adopted an overly deferential posture, and employed an excessively relaxed standard of review for justifying laws that infringe the right to freedom of expression.71 In most of the seminal cases involving freedom of speech, the Court has simply applied what Americans would call a "rational-basis" standard of review- finding that the limitation of the right is justified so long as the court can establish that the law was enacted to achieve a legitimate objective (normally cast in terms of the "public welfare"), and that there is a rational connection or reasonable relationship between that objective and the means adopted to achieve it. 79 Even when the Court has purported to employ a slightly higher standard of review, engaging in some form of "balancing of interests," it has done so in a manner highly deferential to the government interest, and has invariably upheld the challenged law.so In reviewing the seminal cases in which the Supreme Court and lower courts have dismissed constitutional freedom of expression claims, Matsui

ConcernedAbout Declining Media Freedom in Japan, Apr. 11, 2016, https://rsf.org/en/news/rsf- concerned-about-declining-media-freedom-japan; International Federation of Journalists (IFJ). 78. Shigenori Matsui, Freedom of Expression in Japan, 38 OSAKA UNIV. L. REV. 13 (1991) [hereinafter Matsui, Freedom of Expression]; see also P OEJO lik (2014). [MOTOHIRO HASHIMOTO, FREEDOM OF EXPRESSION: THEORY AND INTERPRETATION (2014)]. 79. Matsui, Freedom of Expression, supra note 78, at 21; see also P $ , $U RP)(2005) [NOBUYOSHI ASHIBE, READING CONSTITUTIONAL CASES (2005)], at 184-207. 80. Matsui, Freedom of Expression, supra note 78, at 16-18, citing the Hakata Station TV Film case, Supreme Court of Japan, 23 Keishii 11 (1969), 1490, and the Sarufutsu case, Supreme Court of Japan, Grand Bench, 28 Keishit 9 (1974), 393. 480 HASTINGS CONSTITUTIONAL LAW QUARTERLY [Vol. 45:3 also examined the range of statutes that severely limit both freedom of expression and freedom of association and assembly in Japan, in the form of both content-based regulation and content-neutral controls."' The substance of some of these, such as the Public Offices Election Act, which prohibits the distribution of pamphlets during election campaigns,8 2 and the manner in which other laws and public ordinances have been applied,8 3 Seriously limit political speech and political assembly-activity that is considered to be at the very core of the right to freedom of expression. 4 The upshot of all of this is that the government of Japan is not at all reluctant to enact laws that limit the freedom of expression; nor does it have much reason to fear the prospect of such laws being struck down as unconstitutional by the courts. Consistent with this critique, many scholars and commentators in Japan worry that the passage of hate speech laws, or any legislation that might limit freedom of speech, will provide a cynical government with more leverage to further muzzle dissent and debate within Japan. 5 Such concerns are not trivial or entirely misplaced. Even as the legislation was being debated, members of the ruling LDP suggested that the new law could be used against members of organizations resisting the construction of a new American naval base in Okinawa-which, quite obviously, would have been entirely inconsistent with its purported purpose.86 All of this having been said, regardless of the possible insincerity of the government's free speech objections, and the judiciary's past failure to protect the right to freedom of expression, lawmakers and scholars should nonetheless be striving to develop hate speech laws that do respect and comply with the constitutional right to freedom of expression. Their concern for whether a proposed hate speech law violates the constitutional right is

81. Matsui, Freedom ofExpression, supra note 78, at 22-23, 29-30, 33-36. 82. Id. at 23-24. For discussion of why such a facially content-neutral limitation on door-to- door political canvassing should, from the American perspective, have profound content- differential effects and may thus be deemed invalid, see CASS SUNSTEIN, DEMOCRACY AND THE PROBLEM OF FREE SPEECH 171 (1995). 83. Matsui, Freedom ofExpression, supra note 78, at 34-35. 84. See, e.g., THOMAS I. EMERSON, THE SYSTEM OF FREEDOM OF EXPRESSION (1970); ARCHIBALD Cox, FREEDOM OF ExPRESSION (1981); Thomas I. Emerson, Toward a General Theory of the FirstAmendment, 72 YALE L.J. 877 (1963); Thomas Scanlon,A Theory ofFreedom of Expression, in R. M. DWORKIN, THE PHILOSOPHY OF LAW (1977); for more discussion of the rationales underlying theories of freedom of expression, see infra note 91. 85. See, e.g., Kotani, supra note 21. 86. Eric Johnston, LDP Lawmaker Suggests Using Hate Speech Bill on Protesters in Okinawa, JAPAN TIMES, Apr. 26, 2016, https://www japantimes.co.jp/news/2016/04/26/national /politics-diplomacy/ldp-lawmaker-suggests-using-hate-speech-bill-protesters-okinawa/#.WmU 3JM-fOQ. Spring 2018] STRIKING THE RIGHT BALANCE 481 thus entirely valid and legitimate. But in my view, they should be seeking to find the right balance between respect for freedom of expression on the one hand, and the right to equality and not to be discriminated against on the other hand, in fashioning a hate speech law that is both effective and also consistent with the Constitution. In thinking about how to do that, it is often helpful to consider how other countries have grappled with the challenge- for it is certainly not a challenge unique to Japan. Unfortunately, however, Japanese lawmakers, jurists, and scholars, most frequently look to the United States for insights on how to approach issues relating to constitutional rights issues.Y' It may also seem quite natural in this instance, given the very robust protection of freedom of expression in the jurisprudence of the First Amendment of the Constitution of the United States. But in the context of hate speech, those robust protections have come at the expense of the rights of minorities. As I have argued elsewhere in more detail, when it comes to considering comparative examples of doctrine and jurisprudence relating to the protection of equality rights, constitutional systems such as that of Canada provide the Japanese with a more fruitful comparison than the United States." And on hate speech too, the United States does not provide a good example.

II. The American Approach The American approach to freedom of expression is quite well known in Japan. It has been invoked both by Japanese scholars arguing more generally for more rigorous protection and enforcement of free speech rights by Japanese courts, and it has been deployed more recently by those resisting the pressure for stricter hate speech laws. But for reasons that I will develop below, I would suggest that while the American example may be helpful for bolstering demands for more robust protections of freedom of expression more generally, it is not the best example for dealing with hate speech, or even for developing particular doctrinal approaches to freedom of expression under the Japanese Constitution. The American system has not recognize a balance between the rights to freedom of speech and the right to equal protection in resolving the hate speech issues, but rather privileges speech over equality in a way that should not be emulated. The Canadian example, I will argue, strikes a more balanced approach. What is more, the Canadian and Japanese systems share some similarities, such as a greater sensitivity to the relevant international law obligations, and even the text of the respective

87. See, e.g., Matsui, Freedom of Expression, and Hashimoto, supra note 78, and Ichkawa and HIGAKI, supra note 34; and more generally see e.g., PERE, WA II: AfMi (1994) [ASHIBE NOBUYOSHI, CONSTITUTIONAL LAW VOL. II: THEORY OF RIGHTS (1994)]. 88. Martin, Glimmers ofHope, supra note 10. 482 HASTINGS CONSTITUTIONAL LAW QUARTERLY [Vol. 45:3 constitutional provisions. Together, these should make the Canadian example an important consideration for Japanese law and policy makers, jurists, and scholars alike.

A. Overview-The Pigeonhole Approach It may be apposite to begin by recalling the basic landscape of the American approach to freedom of speech. The First Amendment provides for a very broad and apparently unqualified right: "Congress shall make no law . . . abridging the freedom of speech, or of the press; or the right of the people to peacefully assemble .. ."8 The Supreme Court has developed a rather complex web of doctrine so as to permit some limits on expression. However, it remains the case that the First Amendment right to freedom of speech is the most powerful of rights among the American individual constitutional rights, and the U.S. courts tend to privilege freedom of speech over other rights that may come into tension with free speech. Analysis in U.S. freedom of speech cases typically revolve around two questions: First, what is the nature of the law or regulation that is alleged to be limiting speech; and second, what is the nature of the speech that the law is attempting to limit. With respect to the first question, the issue is whether the law aims to limit the content of speech. Laws or policies that purport to limit speech on the basis of its content, or worse still its viewpoint, are viewed with the greatest suspicion, and will be held to the strictest scrutiny, or the highest standards of review, for purposes of justification. But this is only so for speech that is fully protected. For in response to the second doctrinal question, regarding the nature of the speech that is subject to limitation, the Supreme Court has carved out a number of categories of speech that have been deemed to be unprotected, or to be lesser protected, by the First Amendment. Thus, if the speech that is limited by the impugned law is determined to fall within one of these categories, even if the law is designed to limit the content of the speech, 90 the law will be subjected to a less strict level of scrutiny-either because such speech is said not to come within the scope of the right of freedom of speech at all, is of "lower value" in terms of the underlying rationales for protecting freedom of speech, 91 or

89. U.S. CONST. amend. I. 90. This phrase requires some qualification, as will be discussed in the analysis of the case R.A. V v. City of St. Paul, 505 U.S. 377 (1992), see infra Part II-C. 91. The three rationales most frequently articulated are that freedom of expression is essential to: further the democratic process; facilitate the search for truth in a "marketplace of ideas"; and allow for the realization of personal actualization. A fourth that is sometimes included, is "promoting tolerance." See, e.g. ERWIN CHEMERINSKY, CONSTITUTIONAL LAW: PRINCIPLES AND Spring 2018] STRIKING THE RIGHT BALANCE 483 because it has been long established that the limitation on such speech can be justified under strict scrutiny and so courts need not go through the process every time the category of speech is implicated. 9 2 Among the unprotected and lesser protected categories of speech are defamation, obscene expression, and child pornography. The most germane to our analysis, however, are "fighting words," "incitement to violence or crime," and "true threats." These are the most relevant to our analysis precisely because no distinct category for "hate speech" has yet been developed by the courts. As a result, legislators seeking to craft (and later defend) laws aimed at constraining hate speech, must do so in a manner that characterizes the speech to be limited in such a way that it can fit within one of the established categories of lesser-protected speech. In other words, if the courts are not satisfied that the speech that is being limited by the challenged law fits into one of the categories for unprotected or lesser-protected speech, then the challenged law will be subject to strict scrutiny as a content-based regulation 93 -that is, the government will have to prove that the law serves some compelling state interest, and that the means chosen to achieve that objective is necessary, or at minimum carefully tailored and the least restrictive alternative.94 It is far less likely to survive a stnict scrutiny standard of review. What is more, as I will discuss below, once there is a finding of content-based limitation, the Court tends to presume it to be fatal and seldom embarks on a rigorous application of the strict scrutiny justification analysis. Most of the work is done on the question of categorization. One of the criticisms of the

POLICIES 973-74 (2015) [hereinafter CHEMERINSKY, PRINCIPLES AND POLICIES]. Yet another rationale that is sometimes distinguished from the argument for democratic process, is a justification from social contract theory-which can be distinguished on the grounds that this rationale would not accept suppression of speech that argues for the overthrow of the democratic system itself. See Michael Rosenfeld, Hate Speech in Constitutional Jurisprudence: A Comparative Analysis, 24 CARDOZO L. REv. 1523 (2003), citing JOHN RAWLS, A THEORY OF JUSTICE 11-12 (1971), and more generally, FREDERICK SCHAUER, FREE SPEECH: A PHILOSOPHICAL ENQUIRY (1982). 92. There are differences of view as to whether, for instance, defamation is "unprotected" as being outside the scope of the right, or whether the justification for its limitation is simply so well entrenched that it need not be repeated. The fact that the Supreme Court narrowed the exception for defamation in the relatively recent case of New York Times Co. v. Sullivan, 376 U.S. 254 (1964), would suggest that it is the latter, as does the reasoning of the majority inR.A. V v. City ofSt. Paul, 505 U.S. 377 (1992), but debate remains. There is, obviously, a massive scholarly literature on the First Amendment and freedom of speech. 93. Turner Broad. Sys. v. FCC, 512 U.S. 622, 640-41 (1994). 94. See, e.g., Brownv. Entm't Merchants Ass'n, 131 S. Ct. 2729, 2738 (2011) (holding that a law limiting violent video games was content based, and thus "invalid unless California can demonstrate that it passes strict scrutiny that is, unless it is justified by a compelling government interest and is narrowly drawnto serve that interest"); and see Bursonv. Freeman, 504 U.S. 191 (1992) (upholding a content based limitation on speech as nonetheless satisfying strict scrutiny). 484 HASTINGS CONSTITUTIONAL LAW QUARTERLY [Vol. 45:3 American approach is that the Supreme Court has at once continued to adhere to this rather rigid and somewhat clumsy framework of pigeonhole categories, but at the same time been exceedingly reluctant to create new categories to address such harmful forms of expression as hate speech.95

B. The Early Cases: Group Libel, Incitement, Fighting Words The Supreme Court of the United States came closest to recognizing some form of lesser protected category for hate speech in the 1952 case of Beauharnaisv. Illinois.96 An accused man challenged the constitutionality of an Illinois criminal law provision that made it unlawful to create or disseminate racist material which "exposes the citizens of any race, color, creed or religion to contempt, derision, or obloquy . .. ."97 He was charged for having circulated a petition calling upon the City Council of Chicago to halt the invasion of white neighborhoods by "the Negro," and arguing that the white race must be united by the "need to prevent the white race from being mongrelized by the negro . .. [and] the aggressions, rapes, robberies, knives, guns and marijuana of the negro." 98 The Court upheld that statute, condemning racism and providing a powerful argument for the importance of laws such as the one in question. But, it upheld the law by suggesting that racist expression was analogous to, and indeed a species of, criminal libel, thus fitting the speech into the well-established category of defamation as an unprotected or lesser protected form of speech. 99 So even in this high-water mark case for upholding a form of hate speech law, the Court tried to somewhat awkwardly shoehorn the law into an existing pigeonhole, rather than create a new category of less protected speech. And because it chose libel as the most convenient category, the judgment was short lived. For while Beauharnaishas never been explicitly overturned, it is widely viewed as being no longer good law. oo This is in part because its assumptions about the scope of the category of less protected defamatory speech were radically modified in later cases (specifically New York Times Co. v. Sullivan).'0 ' Courts have thus had to cast about for other pigeonholes. Subsequent

95. See, e.g., John Paul Stevens, The Freedom of Speech: Address of the Honorable John PaulStevens, 102 YALE L.J. 1293, 1300-12 (1993). 96. Beauharnaisv. Illinois, 343 U.S. 250 (1952). 97. Id. at 251. 98. Id. at 252. 99. Id. at 254-56. 100. Colin v. Smith, 578 F.2d 1197, 1204-05 (7th Cir. 1978); ERWIN CHEMERINSKY, CONSTITUTIONAL LAW 1365 (3rd ed. 2009) [hereinafter, CHEMERINSKY, CONSTITUTIONAL LAW]; Rosenfeld, supra note 91, at 1536. 101. New York Times Co. v. Sullivan, 376 U.S. 254 (1964). Spring 2018] STRIKING THE RIGHT BALANCE 485 decisions have characterized laws limiting racist expression and other hate speech as limitations on speech that falls within such lesser-protected categories as fighting words and incitement. But as we will see, the concerns regarding such lesser-protected speech-the concerns which gave rise to the categories in the first place-are quite different from those that animate governments to limit hate speech. And legislative efforts to limit hate speech do not typically satisfy the criteria for such categories. The courts have exempted the "fighting words" and the "incitement to criminal or violent activity" categories from full First Amendment protection primarily for reasons of public order. More specifically, the concern is that the speech will provoke or elicit a response that will be violent or criminal. To begin with "incitement," the primary concern underlying the exception is with imminent harm, either through the commission of illegal acts or by violence, which is likely to be directly incited by the speech. In the early development of this category, the doctrine required the government to establish that the speech prohibited by the law constituted a "clear and present danger" of bringing about some illegality that the government had a right to prevent.1 02 The doctrine evolved over the following decades and crystallized in its current form in the so-called Brandenburgtest. The case, Brandenburgv. Ohio,10 3 involved the prosecution of a leader of a Ku Klux Klan group for advocating violent action against the state, when he threatened such action if the federal government continued to "suppress the white race." While he and other members of the Klan had been recorded on film making virulently hateful statements about African Americans and Jews, that was not the primary legal grounds for the prosecution. Rather, he was prosecuted under Ohio's "syndicalism" statute, one of many similar state laws that made it an offense to advocate certain forms of violence, sabotage, or terrorism as a means of accomplishing a change in industrial ownership or to effect political reform104-a reflection of the anticommunist paranoia in the immediate aftermath of the Russian Revolution. The Supreme Court overturned earlier cases that had upheld such laws,to holding that the mere advocacy of violence or illegality could not be prohibited, "except where such advocacy is directed to inciting or producing imminent lawless action and is likely to incite or produce such action."106 The focus, therefore, was entirely on the activity that the expression will provoke, and

102. Schenckv. United States, 249 U.S. 47 (1919). 103. Brandenburgv. Ohio, 395 U.S. 444 (1969). 104. Whitney v. California, 274 U.S. 357 (1927). 105. Brandenburg, 395 U.S. at 447-48. 106. Id. 486 HASTINGS CONSTITUTIONAL LAW QUARTERLY [Vol. 45:3 it is only where such reaction will likely be immediate, and violent or unlawful, that the speech may be limited. The Brandenburg test that was thereby established carved out this narrow category of speech that is less protected under the First Amendment: Namely expression that constitutes incitement to imminent lawless or violent action. What becomes clear, therefore, is that Brandenburgwas itself not really about hate speech as a matter of law, even though the accused had certainly engaged in the propagation of hate speech. And equally clear is that efforts to justify hate speech laws on the grounds that they regulate a particular form of "incitement" will likely not satisfy the Brandenburg test, since the primary purpose of hate speech laws is not to prevent a possible violent reaction to the hateful speech. Rather, there are two primary objectives that hate speech laws arguably seek to achieve: first, is to prevent the various kinds of harm that such speech causes to members of the minority group, on an individual and collective basis; and second, is to prevent the injury that the hate speech does to the values of tolerance and equality within the society at large, by fostering hatred and discrimination. 0 7 Some narrow band of hate speech might indeed incite a violent reaction, but that is not the primary concern, and the category of speech is not defined by whether the reaction will be violent or unlawful. The doctrine relating to the less-protected category of "fighting words" does not fare much better as a basis for justifying hate speech laws. Again, the emphasis is almost entirely on whether the offending expression is likely to provoke a retaliation or violent response.os The case that established the category, Chaplinsky v. New Hampshire, actually included a second element, that of expression amounting to the kind of insult that would likely cause harm to the listener. But that element has not been upheld since, and indeed the entire category has been significantly narrowed in subsequent cases.1 09 It is only language likely to provoke an immediate violent response that now comes within the category." 0 But even so, laws purporting to limit fighting words are typically struck down for being excessively vague and overly broad in any event."' A plan by neo-Nazis to march in uniform through a prominently Jewish suburb of Chicago, called the Village of Skokie, gave rise to a series of cases in the late 1970s that are often viewed as exemplifying the thorny

107. For more on the nature of such harm, see the sources in notes 38, and text associated with infra notes 174-181. 108. Chaplinsky v. New Hampshire, 315 U.S. 568 (1942). 109. CHEMERINSKY, CONSTITUTIONAL LAW, supra note 100, at 1347-49. 110. Id. 111. Id. Spring 2018] STRIKING THE RIGHT BALANCE 487 constitutional issues surrounding efforts to limit hate speech. In response to the announced plan by the National Socialist Party of America ("NSPA") to conduct this deliberately provocative march, Skokie enacted a number of ordinances. One of these prohibited "the dissemination of any materials with the Village of Skokie which promotes and incites hatred against persons by reason of their race, national origin, and is intended to so." A second ordinance prohibited public demonstrations by members of political parties while wearing "military-style" uniforms.11 2 The ordinances also required the application for permission, and the acquisition of insurance. The NSPA challenged these ordinances and there were a series of cases in both State and Federal courts. The most important decision in the line of cases, Colin v. Jones, was handed down by the Seventh Circuit Court of Appeals, and held the ordinances to be in violation of the First Amendment freedom of speech clause.11 3 The court first found that the key ordinance, the prohibition on the dissemination of hateful materials, could not be characterized as either constituting a limitation on "fighting words" within the meaning of Chaplinsky, or incitement to violence or unlawful acts, within the meaning of Brandenburg."4 The Village had, indeed, conceded that the ordinance had not been enacted out of concerns over a violent response to the march. Instead, the primary argument of the Village of Skokie was that the ordinance was a limitation on hate speech, which could be justified pursuant to the holding in Beauharnaisv. Illinois. The court, however, rejected this central argument for two reasons. First, because the court read even Beaharnaisas being primarily grounded in a concern that the limited group libel could "cause violence and disorder." That, it pointed out, was indeed the primary justification for the criminal libel laws that the court had invoked in Beauharnais."5 And, again, concern over violent reaction to the march was not at issue in the case at hand. Second, and more importantly, the Court expressed the strong view that Beauharnais was no longer good law in any event, given how the scope of both civil and criminal libel law had been narrowed by the Supreme Court in the intervening years.11 6 So to the extent that Beauharnaismight once have been viewed as a possible foundation for

112. Colinv. Smith, 578 F.2d 1197, 1199-1200 (7th Cir. 1978). 113. Id. at 1199-1200. The Supreme Court declined to review the decision (Blackmun, J. and White, J. dissenting on the grounds that any conflict between the 7th Circuit decision and the judgment in Beauhranaisv. Illinois should be resolved by the Supreme Court, since Beauhranais had neverbeen overturned); cert. denied, Smithy. Collin, 439 U.S. 916, 919 (1978). 114. Colinv. Smith, 578 F.2d 1197, 1203 (7th Cir. 1978). 115. Collin, 578 F.2dat 1204. 116. Id. at 1205. 4889 HASTINGS CONSTITUTIONAL LAW QUARTERLY [Vol. 45:3 justifying hate speech regulation, that hope was put to rest in the Skokie litigation. A final argument made by the Village of Skokie was that the ordinance was aimed at saving residents, many of whom were Holocaust survivors, from the severe emotional distress likely to be caused by the march (emotional distress being a new cause of action made permissible by recent developments in tort law). The Court easily dispatched with that argument, noting that principles grounding civil liability for the intentional infliction of severe mental distress could not possibly justify a criminal statute of general application to expression otherwise protected by the First Amendment." 7 To top it all off, the Court found the ordinance vague and overly broad, in the best tradition of efforts to regulate fighting words or incitement." 8

C. The Recent Cases: Viewpoint Limitation and True Threats Two more recent Supreme Court decisions further exemplify the difficulties with trying to limit hate speech. They do so in part by reflecting the extent to which legislators have tried to craft laws to fit within the established categories of lesser-protected speech; and in part by reinforcing how reluctant the Court is to accept any viewpoint-based limitation on speech, regardless how hateful or harmful such speech may be. The first of these cases ostensibly involved "fighting words." In R.A. V v. City of St. Paul,"9 the issue was a municipal ordinance that prohibited the use of symbols or objects, such as a burning cross or Nazi swastika, "which one knows or has reasonable grounds to know arouses anger, alarm or resentment in others on the basis of race, color, creed, religion or gender." 2 0 It can be inferred that the essential purpose of this law was to limit a species of hate speech, but it was framed in a manner designed to satisfy the "fighting words" test. The law was used to prosecute two young men who had erected a burning cross in the yard of the house of an African-American family. The intent and meaning of the action was unambiguous-the symbolic power of a burning cross in the United States is difficult to overstate. It is a symbol associated with the Ku Klux Klan and its campaign of terror against African Americans in particular, and minorities in general. As the Court stated in a subsequent case, "whether the message is a political one or whether it is also meant to intimidate, the burning of a cross is a 'symbol of hate.'""21

117. Collin at 1206. 118. Id. at 1207. 119. R.A.V. v. City of St. Paul, 505 U.S. 377 (1992). 120. Id. 121. Virginia v. Black, 538 U.S. 343, 344 (2003). Spring 2018] STRIKING THE RIGHT BALANCE 4899 The City of St. Paul had tried to narrow the scope of its prohibition on fighting words (which, as mentioned earlier, are frequently struck down for being vague and overbroad) to only those symbols that had salience due to their relationship with personal characteristics of race, color, creed, religion, and gender. But this effort to narrow the scope of the law was precisely what the Court held that it could not do. Justice Scalia, writing for the majority, found that because the ordinance distinguished between different kinds of fighting words based on content, suppressing hateful expression against certain groups but not others (such as political groups, union members or homosexuals), it was not only a content-based regulation of speech, but was indeed a viewpoint-based limitation constituting "special prohibitions on those speakers who express views on disfavored subjects."l 22 Though fighting words constituted a category of less-protected speech, that did not mean that the category was entirely outside of the scope of the First Amendment right, and thus the government was not entitled to discriminate in its limitation of fighting words.1 23 It is somewhat striking that Justice Scalia was far more concerned about the discrimination against certain forms of hateful speech, than he appeared to be about the discriminatory impact and harm caused to minorities by such hateful speech. Indeed, he scarcely addressed the issue of the harmful consequences of such speech at all. This may be in part because the entire issue had been framed as one of "fighting words." Thus, the concern was over the "reaction" that such words would provoke, rather than on hate speech per se, and the nature and extent of the harm it would cause to the minorities it targeted. St. Paul had tried to get beyond the pure fighting words argument with a claim that even if the ordinance was a content-based regulation, it was nonetheless permissible as a regulation aimed at limiting "secondary effects," pursuant to the Court's decision in City of Renton v. Playtime Theatres, Inc.12 4 But Justice Scalia swatted that argument aside, noting that "listener's reactions to speech are not the type of 'secondary effects' we referred to in Renton . . . the emotive impact of speech on its 2 5 audience is not a 'secondary effect.'" The justification analysis in the judgment included some passing reference to the harm that might be caused by the kind of hateful expression

122. R.A.V, 505 U.S. at 391. 123. Id. at 388. 124. City of Rentonv. Playtime Theatres, Inc., 475 U.S. 41 (1986). 125. R.A.V, 505 U.S. at 395 (citing Boos v. Barry, 485 U.S. 312 (1988) and City ofRenton, 475 U.S. 41) (internal citation and quotation marks excluded). 490 HASTINGS CONSTITUTIONAL LAW QUARTERLY [Vol. 45:3 that St. Paul was trying to constrain, but it was perfunctory at best.1 2 6 Having determined that the ordinance was a viewpoint-based limitation within the category of fighting words, Justice Scalia indicated that it must therefore be subject to strict scrutiny.1 27 Yet that analysis was neither searching nor rigorous, which is part of a pattern in these types of cases. Justice Scalia accepted the City of St. Paul's argument that the purpose of the law was a compelling state interest-namely "to ensure the basic human rights of members of groups that have historically been subjected to discrimination, including the right of such group members to live in peace where they wish." 28 But, in insisting that the law be necessary to achieve that objective, he asserted that "the existence of adequate content-neutral alternatives" was fatal to the justification.1 29 Yet he provided no explanation of what those other alternatives might have been, nor indicated what evidence had been adduced to support the assertions, other than to suggest that a completely undifferentiated prohibition on fighting words would have sufficed.1 30 That dubious proposition, of course, is undermined by the fact that fighting words legislation frequently runs afoul of vagueness and over-breadth principles.131 As I will return to discuss in more detail below, it also does not take seriously the idea that the kind of hate speech that targets certain identifiable groups is particularly pernicious, and that the state thus has a particular interest-an interest informed by constitutional values enshrined in the equal protection clause-in limiting such hate speech. The City of St. Paul clearly understood this idea, as was reflected in choices made in drafting the legislation, but the pigeonhole categories of constitutional doctrine also forced the City to couch the prohibition in the language of "fighting words," and the Court in turn dealt with the law entirely in terms of such "fighting words." The second more recent case that is viewed as being part of the hate speech canon, Virginia v. Black,132 also involved yet another instance of cross burning. The Court in this case considered and purported to apply the holding in R.A. V v. City ofSt. Paul. It is typically considered a "hate speech" and "fighting words" case, and it is read by some as possibly creating some space for hate speech laws. This is because, while the majority stuck down

126. For a detailed analysis of how the courts in the American hate speech cases have tended to discount and even trivialize the harm, see Moran, supra note 38. 127. R.A.V, 505 U.S. at 395. 128. Id. 129. Id. 130. Id. 131. CHEMERINSKY, CONSTITUTIONAL LAW, supra note 100, at 1347-49. 132. Virginia v. Black, 538 U.S. 343 (2003). Spring 2018] STRIKING THE RIGHT BALANCE 491 parts of the challenged law for other reasons, it held that a state could indeed prohibit the burning of crosses for the purposes of intimidation.1 33 But while I applaud such attempts to use the judgment to open up such a space, in my view, the decision does not establish the basis for permitting the kind of hate speech laws that I am discussing here. It does not come close to creating a new lesser-protected category of hate speech, and indeed it viewed the law as coming within yet a different narrow and long-established pigeonhole. This becomes clear with a closer examination of the Court's reasons. At issue was a Virginia law that made it a criminal offence to burn a cross with the intent of intimidating any person or group of persons, on the property of another person or in a public place. The law further provided that the burning of the cross was itself primafacie evidence of the intent to intimidate. The case involved the amalgamation of three separate prosecutions-one against a Ku Klux Klan leader for burning a cross at a Klan rally, and the other two against a group of three individuals who had burned a cross in the yard of a black neighbor. The law had been challenged in all three cases as violating the First Amendment. Writing for the majority, Justice O'Connor reviewed the history of the Ku Klux Klan and the symbolism of cross burning in America and found (as quoted earlier) that the burning of a cross is always a symbol of hate. But the issue was not whether it was an expression of hatred, but whether it was used with an "intent to intimidate." The term was construed very narrowly, as meaning a motivation "to intentionally put a person or a group of persons in fear of bodily harm."l34 While the Klan's history of burning of a cross was often intended to make people fear for their life, the Court held that it was clearly not always used for that purpose. It may always be a symbol of hate, but it may not always be meant to make specific people fear imminent violence. Justice O'Connor went on to write "sometimes the cross burning is a statement of ideology, a symbol of solidarity. It is a ritual used at Klan gatherings, and its used to represent the Klan itself." 35 Hateful, yes, and while she never addressed it, it is even likely to always cause fear, a sense of oppression and denigration on the part of black Americans, but not always intimidation in terms of intending to cause fear of imminent bodily harm. In this sense, while most of the separate opinions in Virginia v. Black discuss R.A. V v. City ofSt. Paul at length, it should not be misunderstood as

133. Indeed, Rory Little, in his article for this symposium, makes this argument. See Rory Little, Hating Hate Speech: Why CurrentFirstAmendment Doctrine Does Not Condemn a Careful Ban, 45 HASTINGS CONST. L.Q. 577 (2018). 134. Virginiav. Black, 538 U.S. 343, 349 (2003). 135. Id. at 366. 492 HASTINGS CONSTITUTIONAL LAW QUARTERLY [Vol. 45:3 being a case about "fighting words." Leaving aside the factual similarity of burning crosses, the Court considered R.A. V v. City ofSt. Paul for purposes of assessing whether the Virginia law was a viewpoint-based limitation within a lesser-protected category of speech. However, the lesser-protected category of speech is that of "true threats," not "fighting words."'36 Justice O'Connor explained that true threats constitute a "serious expression of an intent to commit an act of unlawful violence to a particular individual or group," whether or not the speaker actually intends to carry out the threat; and the prohibition "is aimed at protecting 'individuals from the fear of violence and from the disruption that fear engenders.'" 37 The prohibition on the burning of crosses with intent to intimidate, therefore, was a prohibition on a form of true threat, not a prohibition of fighting words, and not a constraint on hate speech per se. The Court held that the prohibition on cross burning, while a content- based limitation within the proscribable category of "true threats," differed from the viewpoint-based limits in R.A. V v. City of St. Paul, because it was not limited to prohibiting an intent to intimidate any particular group, as defined by race, gender, or any other such personal characteristic. Notwithstanding the very powerful history of the Ku Klux Klan using cross burning to intimidate and terrorize African Americans, Justice O'Connor held that "as a factual matter it is not true that cross burners direct their intimidating conduct solely to racial or religious minorities." 38 As dubious as that claim might seem, in essence the argument was that Virginia could prohibit a particularly virulent form of intimidation, without running afoul of the viewpoint-based limitation problem that the City of St. Paul ordinance had. And it is telling that Justice Scalia, who had written the majority opinion in R.A. V v. City ofSt. Paul, concurred with that part of the decision. 139 In the final result, however, the Court struck down the Virginia law. It did so because of the clause of the provision that made cross burning prima facie evidence of an intent to intimidate. In explaining why and how this presumption rendered the provision unconstitutional, the Court revealed just how narrow its holding was. The presumption meant, Justice O'Connor

136. Black, 538 U.S. at 359. 137. Id. at 359-60 (citing Watts v. United States, 394 U.S. 705, 708 (1969)) (internal quotes omitted). 138. Black, 538 U.S. at 362. 139. Id. at 368 (The concurring opinion of Justices Souter, Kennedy, and Ginsburg, dissented strongly on this point, denying that cross burning merely represented a form of intimidation that could be divorced from its historical ideological message. They also held that the law would run afoul of the R.A. V v. City of St. Paul prohibition on viewpoint-based limits within proscribable categories of speech); id. at 383-84. Spring 2018] STRIKING THE RIGHT BALANCE 493 argued, that the State could "arrest, prosecute and convict a person solely on the fact of cross burning itself," which would "create an unacceptable risk of the suppression of ideas."o40 This was so, she reasoned, because burning a cross might not be for the purpose of "intimidation" but rather could be a "statement of ideology, a symbol of group solidarity .... It is a ritual used at Klan gatherings, and it is sued to represent the Klan itself."' 4' Notwithstanding that the ideology is one of racist white supremacy, that the Klan is considered the oldest domestic terrorist organization in America, and that cross burning is closely associated with arson, beatings, and lynching, this was core political speech protected by the First Amendment. Even more troubling for those looking to the decision in Virginia v. Black as a basis for justifying some narrow form of hate speech laws, the concurring decision of the more "liberal" justices was even more restrictive. Justice Souter, writing for himself, Justice Ginsburg, and Justice Kennedy, held that the prohibition on cross burning with an intent to intimidate was itself an unconstitutional content-based limitation within a lesser-protected category, in violation of the holding in R.A. V v. City of St. Paul.142 While the statute did not explicitly discriminate among different forms of "fighting words" on the basis of race, creed, religion or gender, as the ordinance had in R.A. V v. City of St. Paul, Justice Souter argued that cross burning was clearly an anti-black symbol, and thus it was clear that the law was in effect singling out for prohibition one specific form of intimidation that was tied to a particular ideology.1 43 The state's apparent "discrimination" against the hateful and terrorist expression of the cross burner was, in essence, understood to do greater violence to a constitutional right than the state permitting people to employ the most hateful and terrifying symbol in American history to denigrate, terrorize, and foster prejudice and discrimination against a racial minority.

D. Closing Observations In short, laws that have either been enacted or deployed for the purpose of prohibiting expression that might be considered hate speech, have been routinely struck down. In purely doctrinal terms, this was primarily because the court has resisted any attempt to create a new category of lesser-protected (or "proscribable") speech. Thus, legislators have either sought to craft hate speech laws in such a way as to fall within an existing category, as in R.A. V

140. Black, 538 U.S. at 365. 141. Id. at 365-66. 142. Id. at 380-82. 143. Id. at 384. 494 HASTINGS CONSTITUTIONAL LAW QUARTERLY [Vol. 45:3 v. City ofSt. Paul (fighting words) and Virginia v. Black (true threats), or the speech has been categorized as falling within an existing category by the prosecutors and courts, as in Brandeburg v. Ohio (incitement). Efforts to defend laws as a justified limitation on hate speech as such, have been roundly rejected, as reflected in the Skokie line of cases. What is more, we can make some observations about these categories that are typically implicated in hate speech related cases. For the most part, they are primarily concerned with the effect that the speech may have on the narrow target audience, the persons about whom and towards whom the speech is made. Thus, with "fighting words" the concern is that the speech will provoke the subject and target of the speech into imminent violent reaction. With "incitement," the concern is that the speech will incite the subject to commit imminent violent or criminal acts; and with "true threats," the concern is that the speech will cause the subject to suffer fear of imminent violence. Violence and imminence are key elements of all, and with the exception of those true threats for which there is not intent to actually carry through with the threat, the underlying concern is that the resulting conduct will disturb public order. So there is, ultimately, both a concern about the reaction of the direct subject and the target audience, and (with the exception of true threats) a concern about the secondary effects on society more broadly. As I will return to explain in more detail in Part IV, I would suggest that a narrow category of lesser-protected expression of hate speech would similarly share these characteristics of multi-faceted concern, including the reaction of the target audience, harm to the subject of the speech, and the harm to the broader society. What would be quite different is the absence of violence and imminence as features of the dynamic involved, or as criteria for coming within the category of lesser-protected speech. There are a couple of features almost entirely missing from the American doctrinal approach to hate speech. The first, is a sophisticated understanding or appreciation for the harm that hate speech causes. I will discuss this harm in more detail in both Part III and Part IV, but in summary, the evidence suggests that the harm is threefold. There is serious harm to the individual members of the group who are the subject of the hate speech, in of the form of injury to their sense of self-esteem, a sense of alienation from the community, the fostering of fear and a sense of oppression, and even their exclusion from the public discourse. There is also harm to the broader society in the form of the hatred aroused in the broader society against members of the subject group, and thus the increased likelihood of discrimination and even persecution by the broader population against members of the subject group. Finally, regarding the third aspect of harm, one could argue that the tolerance of this kind of speech, leading to increased hatred and discrimination, does violence to the values and principles that are Spring 2018] STRIKING THE RIGHT BALANCE 495 foundational to a constitutional democracy, and thus ultimately causes harm to the constitutional system itself.44 But the American cases do not at all discuss the harm of hate speech in this manner. Indeed, it should be noted here that while the "subjects" of hate speech are typically the members of the minority group being vilified, the broader population, the non-group majority, is often the target audience. This is entirely lost in the discussion of the issue in the American cases-the entire focus is on the "hurt feelings" of the subjects of hate speech, as though they are the target audience, and the only harm relates to their direct response or reaction to the speech. The effect of the speech on the true target audience, and the broader society, tends to be entirely ignored. The cases tend to trivialize and minimize the harm caused to the victims of hate speech and overlook altogether the extent to which it may stoke hatred in the wider population and thereby contribute to discrimination or worse against the targeted minority groups. 145 The second feature of the American approach to the issue of hate speech, is the remarkable absence of any discussion of a possible relationship between hate speech laws and the constitutional right to equal protection in the Fourteenth Amendment. As I indicated earlier, and will return to in more detail below, the Canadian and international law approaches (among others) view hate speech laws as fulfilling the constitutional right to equal protection from discrimination, thus creating a tension between the two fundamental rights of freedom of expression on the one hand, and on the other hand the right to be treated equally, and not to be discriminated against on the basis of shared personal characteristics tied to identity. As we will see, it is the kinds of prohibited grounds of discrimination in the equality rights in the CanadianCharter ofRights andFreedoms, and in the ICCPR and the CERD, that can most logically inform the scope of hate speech laws.1 46 One could have argued that the viewpoint-based limitations on "fighting words" in R.A. V v. City ofSt. Paul, which were based on "race, color, creed, religion, or gender," were indeed informed by the suspect and quasi-suspect classifications in equal protection jurisprudence. 47 In other words, the viewpoint-based limits were designed to protect persons from the kind of

144. For more on this, see supra notes 38 and infra notes 174-181 and accompanying text. 145. This point is made in some detail by Moran, supra note 38, at 1439-41, 1446-50, 1454- 55. See also Rosenfeld, supra note 91, at 1558 ("the American approach tends to remain blind to the considerable potential harm that hate speech can cause to the equality and dignity concerns of its victims or the attitudes and beliefs of non-target audiences") and 1560. 146. This argument is developed further infra Part IV. 147. Sunstein makes a similar point about Beauharnais,and about hate speech more generally; SUNSTEIN, supra note 82, at 173, 184-85. 496 HASTINGS CONSTITUTIONAL LAW QUARTERLY [Vol. 45:3 discrimination (albeit in the private sphere) that the equal protection clause is supposed to protect against (in the public realm), and that the justification for such limits could have been framed in equal protection terms. But that was not only not raised in the case itself, but it is not typically part of the scholarly discourse on the issue either.148 In essence, in both case law and scholarship, there is little recognition that efforts to limit hate speech implicate a tension between the right to freedom of speech, and the right to be treated equally and not to be discriminated against in society, far less any attempt to find ways to resolve such a tension. Rosenfeld has summed it up well:

In terms of its assumptions, the American approach either underestimates the potential for harm of hate speech that is short of incitement to violence, or it overestimates the potential for rational deliberation as a means to neutralize calls to hate. In terms of impact, given its long history of racial tensions, it is surprising that the United States does not exhibit greater concern for the injuries to security, dignity, autonomy and well being which officially tolerated hate speech causes to its black minority. Likewise, America's hate speech approach seems to unduly discount the pernicious impact that racist speech may have on lingering dormant sentiments still harbored by a non-negligible segment of the white population.1 49

As I will return to discuss in Part IV, if the courts and lawmakers recognized a relationship between hate speech law and the values and interests protected by the equal protection clause, then hate speech laws would look quite different, and would indeed be more narrowly drawn; and the approach to their justification would be different. I will argue below that such laws could be upheld even under the current American doctrinal approach.

III. The Canadian Approach In contrast to the American experience, Canada has seen the enactment of hate speech laws at both the federal and the provincial level, in both criminal law and in human rights statutes, and they have largely survived numerous constitutional challenges. In this section I will introduce the statutory framework in some detail, given that I am suggesting that an exception along the lines developed in Canada might serve as a good example for other systems. I will then explain how the Supreme Court of Canada has justified upholding these laws, and in particular how it has

148. See, e.g., Elena Kagan, Regulation ofHate Speech and PornographyAfter R.A. V., 60 U. CHI. L. REv. 873 (1993). 149. Rosenfeld, supra note 91, at 1559-60. Spring 2018] STRIKING THE RIGHT BALANCE 497 understood such hate speech laws as striking a balance between the right to free expression and the right to equality.

A. The Statutory Framework As indicated above, there is both a criminal law prohibition on the dissemination of hate speech, as well as prohibitions in some (but not all) provincial human rights codes. There was also a prohibition in the Federal Canadian Human Rights Act,s0 but that was rescinded in 2013.'st The criminal law prohibition is obviously the more serious limitation on free speech. However, it is very narrowly drawn and provides for robust defenses, so as to impinge on the right to freedom of expression to the minimum extent possible, while still protecting the rights of identifiable groups typically subject to discrimination and hatred. It is worth considering certain aspects of the Criminal Code of Canadaprovision in detail. It provides that:

Public Incitement of Hatred 319 (1) ... (2) Every one who, by communicating statements, other than in pnivate conversation, wilfully promotes hatred against any identifiable group is guilty of (a) an indictable offence and is liable to imprisonment for a term not exceeding two years; or (b) an offence punishable on summary conviction.

Defences (3) No person shall be convicted of an offence under subsection (2) (a) if he establishes that the statements communicated were true; (b) if, in good faith, the person expressed or attempted to establish by an argument an opinion on a religious subject or an opinion based on a belief in a religious text; (c) if the statements were relevant to any subject of public interest, the discussion of which was for the public benefit, and if on reasonable grounds he believed them to be true; or (d) if, in good faith, he intended to point out, for the purpose of removal, matters producing or tending to produce feelings of hatred toward an identifiable group in Canada.

150. Canadian Human Rights Act, R.S.C. 1985, c. H-6. 151. Section 13(1) of the Canadian Human Rights Act was repealed June 26, 2013. 4989 HASTINGS CONSTITUTIONAL LAW QUARTERLY [Vol. 45:3 Definitions (7) In this section, communicating includes communicating by telephone, broadcasting or other audible or visible means; (communiquer) identifiable group has the same meaning as in section 318 [which provides that identifiable group means any section of the public distinguished by colour race, religion, national or ethnic origin, age, sex. sexual orientation, gender identity or expression, or mental or physical disability] (groupe identifiable).152 public place includes any place to which the public have access as of right or by invitation, express or implied; (endroitpublic) statements includes words spoken or written or recorded electronically or electro-magnetically or otherwise, and gestures, signs or other visible representations. (declarations).153

The first subsection of Section 319, which I have omitted above, prohibits the communication of statements that would incite such hatred as is likely to lead to a "breach of the peace." This is similar to the kinds of speech that would constitute "fighting words" or "incitement" in the American jurisprudence. But Section 319(2) provides for the prohibition of what is really at issue here-communication that is likely to incite hatred and discrimination against identifiable groups, but which does not constitute either incitement of imminent violence or criminal activity, nor is necessarily going to provoke a reaction constituting a disruption of public order. The purpose is more directly to prevent communication that will cause harm to members of identifiable groups, typically minorities, by fostering prejudice, hatred, and discrimination against them. And as we will see, it prohibits speech that is based on the same grounds of discrimination that are at the heart of the equality right that individuals have in their relationship with government, enshrined in Section 15 of the CharterofRights and Freedoms.15 4 The provisions in the various provincial human rights codes tend to be somewhat broader in reach, and do not contain the same level of defenses and qualifications as the Criminal Code provision, for which reason they have been the subject of considerable controversy and criticism, and judicial review. It was partly in response to such criticism that the government under Conservative Prime Minister Stephen Harper repealed the hate speech

152. Criminal Code of Canada, R.S.C., 1985 c. C-46, (last amended by S.C. 2017, c. 13), s. 318. 153. Id. at s. 319. 154. Canadian Charter of Rights and Freedoms, Part I of the Constitution Act, 1982, being Sched. B to the Canada Act 1982, ch. 11, section 15(1) (U.K.) [hereinafter Charter of Rights and Freedoms]. Spring 2018] STRIKING THE RIGHT BALANCE 499 provisions of the CanadianHuman Rights Act.15 5 The human rights codes in the provinces of Canada, and at the federal level, provide a framework of human rights, particularly equality rights, that individuals are to enjoy in their private relations, such as in the context of employment, and access to services. But in addition to the substantive rights, the legislation creates an institutional apparatus for receiving individual complaints of rights violations. They are assessed by the Human Rights Commission, and if found credible, are advanced and advocated by the Commission before an independent Human Rights Tribunal. However, the Tribunal is limited to providing a civil remedy sounding in damages or declarations-and thus the sanctions for violating the hate speech provisions in a human rights code are significantly less serious than a violation of the Criminal Code provision.1 5 6 I will below discuss a recent and important Supreme Court decision upholding the hate speech provisions of the Saskatchewan Human Rights Code, and so I will leave detailed analysis of the language of such provisions until then.1 7

B. Overview of Charter Rights at Issue It is perhaps helpful to introduce the relevant constitutional landscape under the Canadian CharterofRights and Freedoms, before launching into how particular hate speech laws are reviewed by the courts. The natural place to begin is with freedom of expression. This is provided for in Section 2(b) of the Charter: "Everyone has the following fundamental freedoms: (b) freedom of thought, belief, opinion, and expression, including freedom of the press and other media communication." 5 s The scope of the right to freedom of expression has been interpreted in a broad and purposive manner, as protecting any nonviolent activity that conveys or attempts to convey meaning or expressive content.1 5 9 Moreover, in assessing whether the challenged government action has limited or infringed this right, the inquiry is again very broad, requiring the court to determine whether the purpose or effect of the government action restricted the freedom.1 60

155. R.S.C., 1985, c. H-6 (as amended). 156. For a succinct explanation of the Canadian Human Rights Act framework, see Julian Walker, Canadian Anti-Hate Laws and Freedom of Expression, LIBRARY OF PARLIAMENT, PUBLICATION NO. 2010-31-E, Mar. 27, 2013, https://lop.parl.ca/Content/LOP/ResearchPublic ations/2010-3 1-e.pdf. 157. See infra Part III-C. 158. Charter of Rights and Freedoms, section 2(b). 159. Irwin Toy v. Quebec [1989] 1 S.C.R. 927, 969-70 (Can.). 160. Id. 500 HASTINGS CONSTITUTIONAL LAW QUARTERLY [Vol. 45:3 Quite unlike the American doctrinal approach of applying different levels of scrutiny depending on the kind of regulation at issue, or the category of speech that has been limited, the Supreme Court of Canada has developed a single doctrinal framework for the justification stage of all Charter analysis. This framework is indeed suggested by the language of Section 1 of the Charter, which provides that the Charter "guarantees the rights and freedoms set out in in it subject only to such reasonable limits prescribed by law as can be demonstrably justified in a free and democratic society."l 6 ' Upon finding that a Charter right has been limited or infringed by government action, therefore, the court moves to determine whether the limitation can be so justified. The analytical framework employed is commonly referred to as the Oakes test, after the case in which it was first elaborated.1 6 2 The Oakes test is quite typical of the proportionality analysis employed in other constitutional systems, and is somewhat similar to the American strict scrutiny test, but with an additional step that requires an explicit analysis of the proportionality between the benefit to be achieved by the law in relation to the harm caused by its violation of the right.1 6 3 In short, it requires that the government first establish that the limitation is prescribed by law; second, that it serves a pressing and substantial objective that is consistent with the values of a free and democratic society; and third, that there is proportionality between the means adopted and the objective to be achieved by the legislation. This third element in turn requires that there is a rational connection between the means and the end, that the means impairs the right in question as little as possible, and that the benefit to be derived from achieving the objective is proportionate to the harm caused by violating the right.1 64 While there has been vigorous and ongoing debate within Canadian constitutional circles over how and to what extent the Oakes test has evolved over time, and how elastic it may be depending on the kind of interests involved,1 6 5 it is reasonably constant when compared to the American differentiated levels of scrutiny. Finally, a few words of introduction are necessary for the equality rights under the Charter, given that the Supreme Court has made explicit reference

161. Charter of Rights and Freedoms, Section 1. 162. R. v. Oakes, [1986] 1 S.C.R. 103 (Can.). 163. See DAVID BEATTY, THE ULTIMATE RULE OF LAW (2004) for a comparative study of proportionality tests. 164. See R. v. Oakes, [1986] 1 S.C.R. 103 (Can.). 165. See, e.g., Sujit Choudhry, So What is the Real Legacy of Oakes? Two Decades of ProportionalityAnalysis Under the Canadian Charter'sSection 1, 35 SUP. CT. REV. 501 (2006); PETER HOGG, CONSTITUTIONAL LAW OF CANADA, Chap. 38 (5th ed. 2012). Spring 2018] STRIKING THE RIGHT BALANCE 501 to these rights in its consideration of hate speech cases. Perhaps of particular interest for Japanese readers, Section 15 of the Charter bears some resemblance to the equality rights enshrined in Article 14 of the Constitution of Japan, in that it explicitly prohibits discrimination and provides a nonexclusive list of grounds of discrimination that are considered prima facie unjust.1 6 6 Specifically, it provides: "15(1) - Every individual is equal before and under the law and has the right to equal protection and equal benefit of the law without discrimination and, in particular, without discrimination based on race, national or ethnic origin, colour, religion, sex, age or mental or physical disability. "167 In contrast to the expansive reading of the right to freedom of expression in Section 2(b), the scope of Section 15 has been interpreted to be somewhat more limited, and does not leave all the work to be done in the justification analysis.1 68 Section 15 obviously governs the relationship between the individual and the state, and protects individuals from discrimination at the hands of the state, and so one might wonder why it is relevant at all to a discussion of laws prohibiting hate speech uttered by private persons. But the courts have recognized that hate speech laws, which serve the purpose of providing individuals with some protection from discrimination within the private sphere, help to fulfill the guarantee of the constitutional right to equal protection and equal benefit of the law. The courts quite explicitly acknowledge that hate speech laws therefore trigger a tension between two countervailing constitutional rights: freedom of speech on the one hand, and the right to be treated equally and as an equal on the other. The right to equality may be attenuated and indirect in this context, in that hate speech laws are primarily aimed at preventing discrimination by private entities, and thus do not directly implicate the equality rights of Section 15. But, in contrast to the American and Japanese approaches, such laws are nonetheless understood as being for the purpose of fulfilling the equal protection of the law provided for in the Charter.1 69

166. See Martin, Glimmers ofHope, supra note 10, for my more detailed comparative analysis of Canadian, American, and Japanese analytical approaches to equality rights. 167. Charter of Rights and Freedoms, s. 15(1). 168. The seminal cases on the interpretation of Section 15 are Law Society of British Columbia v. Andrews, [1989] 1 S.C.R. 143 (Can.) and Law v. Canada (Minister of Employment and Immigration), [1999] 1 S.C.R. 497 (Can.). 169. See infra Part IV-A for more detailed discussion of this point. 502 HASTINGS CONSTITUTIONAL LAW QUARTERLY [Vol. 45:3 C. Upholding the Criminal Prohibition Section 319 of the Criminal Code has been challenged several times as constituting a facial violation of the right to freedom of expression as provided for in Section 2(b) of the Charter. In its first and seminal judgment on the issue, in R. v. Keegstra,170 the Supreme Court of Canada upheld the constitutionality of the limits on freedom of expression created by Section 319. The case involved a high school teacher in Alberta who repeatedly made virulently anti-Jewish statements to his pupils.' 7' He was prosecuted under Section 319(2) of the Criminal Code,1 72 and he challenged the constitutionality of the law. The Supreme Court's analysis in Keegstra provides a framework for thinking about how such hate speech laws can be justified, notwithstanding the limits they impose on freedom of expression, and illustrates a marked departure from the approach taken by U.S. courts. As indicated above, the initial inquiry in freedom of expression cases is to determine whether the conduct that is said to have been limited constitutes activity that conveys or attempts to convey meaning. On that analysis, the expression of Keegstra clearly came within the scope of the protection afforded by the right in Section 2(b). Similarly, the prohibition in Section 319(2) of the Criminal Code just as clearly impinged upon that right, in that it had both the purpose and effect of limiting the kind of expression in which Keegstra had been engaged.1 73 The question for the Court, therefore, was whether the Criminal Code limitation could be justified in accordance with Section 1 of the Charter. In applying the Oakes test in Keegstra, Chief Justice Dickson, writing for the majority, took considerable time to examine in detail the importance of the objective of Section 319(2). There was extensive evidence provided to the court on the incidence of hate propaganda in Canada, and the harmful effects such hate speech had on the individual members of identifiable minority groups. In particular, the evidence suggested that hate speech operated to foster a sense of rejection, alienation and loss of place within the community, and to undermine the individual's sense of self-worth and status in society.174

170. R. v. Keegstra, [1990] 3 S.C.R. 697 (Can.). 171. Id. at 713-14. 172. Criminal Code of Canada, R.S.C., 1985, c. C-46 (as amended), s.319(2). 173. It should be noted that the government tried to argue that hate speech, like violent action, was outside of the scope of the right altogether, but this argument was rejected by the Court. 174. Keegstra, [1990] 3 S.C.R. at 745; The Courtin particular cited the findings of The Special Committee on Hate Propaganda in Canada (the Cohen Committee), which were published in the Report of the Special Committee on Hate Propagandain Canada, 1966, id. at 274-25, 745; A report of the House of Commons Special Committee on Participation of Visible Minorities in Spring 2018] STRIKING THE RIGHT BALANCE 503 The social science evidence also demonstrated that such hate speech has a harmful impact on the society as a whole, in that it can skew attitudes and beliefs, such that the hateful views may gain credence and help foster discrimination and even violence against the target groups.175 Further, it can have an even more fundamental or foundational harm to society, in that it can undermine the very values of tolerance, equality, and respect for human dignity that are essential to liberal democracy.1 76 Citing the experience of Germany with the rise of the Nazi party and the increasing discrimination against Jews in the 1930s, facilitated in large measure by Nazi-sponsored hate propaganda, the Court expressed skepticism that society can always rely on the unfettered and entirely unregulated marketplace of ideas to ensure that the truth will emerge triumphant, at least in the short to medium term-and in the interim, hate propaganda can cause significant harm, and emasculate the principles that form the fundamental fabric of constitutional democracy. 17 Quoting the report of The Special Committee on Hate Propagandain Canada,the Court held that:

We are less confident in the 20th century that the critical faculties of individuals will be brought to bear on the speech and writing which is directed at them . ... While holding that over the long run, the human mind is repelled by blatant falsehood and seeks good, it is too often true, in the short run, that emotion displaces reason and individuals perversely reject the demonstrations of truth put before them and forsake the good they know. The success of modem advertising, the triumphs of impudent propaganda such as Hitler's, have qualified sharply our belief in the rationality of man. We know that under strain and pressure in times of irritation and frustration, the individual is swayed and even swept away by hysterical, emotional appeals. We act irresponsibly if we ignore the way in which emotion can drive reason from the field. 7 8

The Court went on to note evidence that hate speech can have this effect in often subtle and insidious ways, in that even when the message may be

Canadian Society, entitled Equality Now, id. at 745; ReportArising Out ofthe Activities ofthe Ku Klux Klan in British Columbia, John D. McAlpine, 1981, id. at 749; The Canadian Bar Association's Report of the Special Committee on Racial and Religious Hatred, 1984, id. at 749; Law Reform Commission of Canada report, Hate Propaganda,1986, id. at 749. 175. Keegstra, [1990] 3 S.C.R. at 766. 176. Id. 177. Id. at 766-67; id. at 748-49. 178. Keegstra, [1990] 3 S.C.R at 747 (citing the Report of the Special Committee on Hate Propaganda in Canada, 1966). 504 HASTINGS CONSTITUTIONAL LAW QUARTERLY [Vol. 45:3 consciously rejected by recipients, the underlying premises, of racial or religious inferiority, can persist and affect attitudes and behavior.1 79 The Court also made the important point, one often missed in more recent debates about the extent to which hate speech laws stifle free expression, that hate speech can itself operate to limit the right to freedom of expression and undermine the free marketplace of ideas. This is so because hate speech, in the form of extreme forms of speech aimed at vilifying, denigrating, and even dehumanizing members of minority groups within society, not only distorts the search for truth, but suppress and silence the voices of the members of the target minority.'s0 Members of the hated group are effectively muzzled and driven from the public arena and fora of debate, and so the marketplace of ideas is entirely deprived of their perspective.' 8 Of considerable significance for those reflecting on possible models, particularly for law and policy makers in Japan, the majority in Keegstra explicitly reviewed both the American jurisprudence, as well as the international law obligations of Canada. The majority conducted a detailed review of the important American cases, and the scholarly treatment of that jurisprudence, but concluded that the Charter required a different balance among the competing rights. 8 2 The Court went on later to explicitly examine how other provisions of the Charter might inform the analysis of the importance of the objective of Section 319(2). In sharp contrast to the Amencan approach, the Court engaged in a very self-conscious balancing of the different rights implicated by the conflict between freedom of expression and hate speech laws. In particular, it considered how the hate-speech provision furthered the right to equality and not to be discriminated against, enshrined in Section 15 of the Charter. It went on to explain how the right to equality was not merely a prohibition on discriminatory laws and other state action, but rather imposed on government an obligation to take action to protect people from discrimination in the private sphere. 8 3 Therefore, the Court reasoned that it was not limited to only considering equality rights in the context the direct relationship between the state and the individual, but rather it could also consider whether laws were designed to reduce private sector discrimination in accordance with this obligation to realize the promise of Section 15.184

179. Id. 180. Keegstra, [1990] 3 S.CR at 762-63. 181. This point is similarly made by Rosenfeld, supra note 91, at 1561-63. 182. Keegstra, [1990] 3 S.CR at 738-744. 183. Keegstra, [1990] 3 S.C.R. at 755 (citing Law Society of British Columbia v. Andrew, [1989] 1 S.C.R. 143, 171 (Can.) (per McIntyre, J.)). 184. Id. at 751-54. Spring 2018] STRIKING THE RIGHT BALANCE 505 The second move that the Court made that is significant in terms of considering the Japanese situation, was to argue that the international law obligations, imposed by the ICCPR and the CERD in particular, were relevant to the justification of the hate speech law. It held that the human rights obligations were not only relevant to interpreting the competing Charter rights in question, but also to considering the importance of the objectives of the hate speech laws in a free and democratic society. What is more, the Court emphasized that the international law treaties themselves contemplate a balance being struck between freedom of expression and the right to equality in imposing an obligation to implement hate speech laws.s5 In applying the proportionality test that is required by Canadian justification analysis in Charter cases, the Court noted that hate speech and discriminatory propaganda tends to be quite far from the core principles that provide the fundamental rationale for the right to freedom of expression. The Supreme Court has accepted the three rationales that are most widely cited as explaining the necessity for freedom of expression, namely: (i) for the functioning of a democracy, and thus political speech in particular should be protected; (ii) for the search for truth in a free exchange of ideas; and (iii) for individual self-actualization and flourishing. 18 6 The Court in Keegstra found that the kind of narrowly defined hate speech prohibited by Section 319(2)- which was deeply hurtful and damaging to the target group members, misleading to listeners within society, and antithetical to tolerance, understanding, and equality within society-was quite distant from these core values.s7 Indeed, as indicated above, the Court argued that hate-speech can hamper and undermine freedom of speech and interfere with the free marketplace of ideas.'8s In sum, the Court concluded that while hate speech is often "political" in some sense, thus placing it in a category that is viewed as most protected by freedom of speech, the very narrowly defined expression in question is in fact antithetical to the fundamental values of a democracy. 89 Finally, it is important to note that the Court carefully examined the definition of prohibited speech in Section 319(2), and particularly analyzed the qualifications and defenses provided for in the provision to ensure that only a very narrow category of expression, uttered willfully for a particular purpose, is captured. We need not delve deeply into the details here, but the

185. Id.. 186. Irwin Toy v. Quebec, [1989] 1 S.C.R. 927, 970-71 (Can.); see also supra note 91 (discussing these rationales). 187. Keegstra, [1990] 3 S.C.R at 761. 188. Id. at 762-63. 189. Keegstra, [1990] 3 S.C.R at 763. 506 HASTINGS CONSTITUTIONAL LAW QUARTERLY [Vol. 45:3 Court focused on several key elements of the law, which required that the expression be public, that it was for the "wilfull" promotion of hatred, and that it is limited to expression that "promotes" (not merely encourages or expresses) hatred, and finally that it is limited to expression targeting a clearly defined "identifiable group." All of this very much narrowed the category of speech included in the prohibition. Moreover, the defenses provided were robust, helping to ensure that only those who were willfully communicating false statements designed to foster hatred could be prosecuted.' 90 Such was the majority of the Court's justification in the seminal case on the issue. Keegstra's conviction was upheld, though in the final analysis it is interesting to note that he received a suspended sentence and a small fine.191 The law and its application has been challenged many times since. The Supreme Court has continued to uphold its constitutionality.

D. Justifying the Human Rights Code Provisions As mentioned at the beginning of this Part, human rights legislation at both the provincial and the federal level has included prohibitions on hate speech that are broader in scope than Section 319 of the CriminalCode, and the violation of which give rise to civil remedies. These provisions too have been challenged as violating Section 2(b) of the Charter, and have been similarly upheld by the Supreme Court. The two most important cases are Canada (Human Rights Commission) v. Taylor,192 a case from 1990 (the same year as Keegstra), in which the Court considered a challenge to hate speech provisions of the Canadian Human Rights Act, and Whatcott v. SaskatchewanHuman Rights Tribunal,193 a much more recent case, in which the Court upheld the hate speech provisions in the Saskatchewan Human Rights Code. At issue in Whatcott was the dissemination of pamphlets that vilified homosexuals and advocated for their being barred from public schools.1 94 The provision at issue in the Saskatchewan legislation, Section 14(1)(b), prohibited the publication, distribution or display of any representation "that exposes or tends to expose to hatred, ridicules, belittles or otherwise affronts the dignity of any person or class of persons on the basis of a prohibited ground." 95 Prohibited grounds of discrimination are listed in the legislation,

190. Id. at 771-81. 191. Peter Bowal & Craig Graham, Whatever Happened to ... Jim Keegstra, LAWNOW (July/Aug. 2012). 192. Canada (Human Rights Commission) v. Taylor, [1990] 3 S.C.R. 892 (Can.). 193. Whatcott v. Saskatchewan Human Rights Tribunal, [2013] 1 S.C.R. 467 (Can.). 194. Id. at para. 8. 195. SaskatchewanHumanRights Code, S.S. 1979, c. S-24.1 (as amended), s. 14(1)(b) (Can.). Spring 2018] STRIKING THE RIGHT BALANCE 507 and these reflect the same prohibited grounds of discrimination that are provided for in Section 15 of the CharterofRights and Freedoms, a point to which I will return. But for now, it will be noted how much broader and looser the scope of the language in the provision is, as compared to Section 319 of the Criminal Code. As might be expected, this language was challenged in the case as being vague and overbroad, and as capturing expression that was well outside the scope of true hate speech. Justice Rothstein, writing for a unanimous Court, examined in some detail the question of what precisely constitutes "hatred," as that term is used in this and similar provisions. In doing so, he referred at length to the Court's decision in Canada (Human Rights Commission) v. Taylor, in which Dickson C.J.C had held that the term "hatred and contempt" had to be interpreted narrowly, in a manner informed by Parliament's objective of protecting the equality and dignity of all individuals. 19 6 The term referred to "unusually strong and deep-felt emotions of detestation, calumny and vilification."l9 7 But Rothstein outlined all the objections that had been mounted against this definition as established in Taylor, including arguments that it was highly subjective, arbitrary, vague, and overbroad. Beginning with the issue of subjectivity, he noted that the application of the provision is based on a reasonable person standard. The determination of whether some publication falls within the scope of the provision, does not depend on the subjective views of the publisher or the victim of the hate speech, but rather on an objective application of the test.1 98 The question posed is whether, "when considered objectively by a reasonable person aware of the relevant context and circumstances, the speech in question would be understood as exposing or tending to expose members of the target group to hatred." 99 As for the word "hatred" itself, and whether it was inherently too emotional to be susceptible to such an analysis, and would thus lead to both subjective and arbitrary results, Justice Rothstein argued that courts are entirely capable of adhering to both the proper meaning of the words employed by the legislature, and applying them in a manner consistent with the legislative objective. 200 He further noted that it is important that courts understand the depth and strength of the word "hatred"," which involves detestation, vilification, and in the context of human rights legislation,

196. Whatcott, [2013] 1 S.C.R. 467 at para. 24. 197. Id. (citing Taylor, 3 S.C.R. 892 at 928). 198. Id. atpara. 35. 199. Whatcott, [2013] 1 S.C.R. 467, para. 35. 200. Id. at para. 38. 5089 HASTINGS CONSTITUTIONAL LAW QUARTERLY [Vol. 45:3 typically includes a component of viewing members of the target group as being inferior. The act of vilification usually involves suggestions of vile characteristics, inherent and immoral deficiencies, and is intended to delegitimize members of the target group as unworthy of respect and inclusion 20 within the broader community. 1 One of the more extreme forms of such vilification, that is common to hate propaganda, is to dehumanize members of the group, typically through the use of animal labels and metaphors.202 The depth and intensity of what is at issue is further informed by considering the legislative objective. Here, the Court again rejected the idea that hate speech law is concerned with mere "hurt feelings" or "humiliation" on the part of the target group, or the expression of merely offensive and repugnant ideas. "It does not, for example, prohibit expression which debates the merits of reducing the rights of vulnerable groups in society. It only restricts the use of expression exposing them to hatred as part of that debate."203 Indeed, delving further into the legislative objective when considering the importance of the "pressing and substantial objective" for purposes of the Oakes test, the Court emphasized that the focus is not on the nature of the ideas at all, but rather the discriminatory effects of the expression, and how those effects constitute the harms discussed above in Keegstra.204 It emphasized, again, that "[h]ate speech is, at its core, an effort to marginalize individuals based on their membership in a group. Using expression that exposes the group to hatred, hate speech seeks to delegitimize group members in the eyes of the majority, reducing their standing and acceptance within society. When people are vilified as blameworthy or undeserving, it is easier to justify discriminatory treatment." 205 It went on to say that the examples of Fascist Italy and Nazi Germany, relied upon by the Court in Keegstra, were not isolated instances, pointing to the former Yugoslavia, Cambodia, Rwanda, Darfur, and Uganda as more modem instances of ethnic cleansing and genocide preceded by rampant hate speech.206 In conducting the proportionality analysis required under the Oakes test, however, the Court did find that one clause of the provision in the

201. Id. at para. 42-43. 202. Id. at para. 45. 203. Id. at para. 51. 204. Whatcott, [2013] 1 S.C.R. 467, para. 73; see also sources inandtext associated with supra notes 38 and 174-181. 205. Whatcott, [2013] 1 S.C.R. 467, para. 71. 206. Id. at para. 72. For the Supreme Court's consideration of hate speech as a factor in the Rwanda genocide, see Mugesera v. Canada (Minister of Citizenship and Immigration), [2005] 2 S.C.R. 100 (Can.). Spring 2018] STRIKING THE RIGHT BALANCE 509 Saskatchewan Human Rights Code was not rationally connected to the objective, and was overbroad and vague.207 It will be recalled that Section 14(1)(b) prohibited the publication or display of any representation "that expose or tends to expose to hatred, ridicules, belittles or otherwise affronts the dignity of any person of class of persons on the basis of a prohibited ground." 2 08 Part of the argument in the case, and indeed much of the broader criticism of the provision, was that the clause "ridicules, belittles or otherwise affronts the dignity of any person. . ." was far too low and vague a threshold, and as such constituted an unjustifiable limitation on free expression. On this point the Court agreed, finding that this clause did "not rise to the level of ardent and extreme feelings" that are essential to the constitutionality of the hate speech limitations.209 It went on to explain that expression that belittles a minority group or "attacks its dignity through jokes, ridicule or insults may be hurtful and offensive" but that "offensive ideas are not sufficient to ground a justification for infringing on freedom of expression. "210 Thus, the Court found this clause not rationally connected to the legislative objective of addressing systemic discrimination, and finding that the provision could survive with this clause having been severed, struck it down.2 1 ' The Court went on to find that the surviving narrow provision was sufficiently carefully tailored, minimally impairing the right, and that the benefits to be obtained by achieving the objective were proportionate to the harm caused by the limited infringement of the right.2 12 In doing so, however, it considered several more important points, which again reflect significant differences when compared to the American approach. In considering the overbreadth question, as part of the minimal impairment element of the analysis, the Court noted that the nature of the expression is a relevant consideration. Not in the creation of specific categories of lesser protected speech as in the American approach, but in considering the extent to which the expression that is limited by the legislation is closer to or further from the core values underlying the right. These core values are, of course, informed by the three rationales that were discussed earlier, namely, furthering the democratic process, the search for truth, or personal fulfillment.2 13 To the extent the expression in question is further away from

207. Whatcott, [2013] 1 S.C.R. 467, paras. 92-94. 208. SaskatchewanHumanRights Code, S.S. 1979, c. S-24.1 (as amended), s. 14(1)(b) (Can.). 209. Whatcott, [2013] 1 S.C.R. 467, para. 89, 111. 210. Id. at para. 90. 211. Id. at paras. 92-94. 212. Whatcott, [2013] 1 S.C.R. 467, paras. 101-147. 213. Id. atpara. 112. 510 HASTINGS CONSTITUTIONAL LAW QUARTERLY [Vol. 45:3 the core values, it will "affect its value relative to other Charter rights, the exercise or protection of which may infringe freedom of expression."2 14 As had been found in both Keegstra and Taylor before this, the Court noted that hate speech "strays some distance from the spirit of s. 2(b), and hence conclude that restrictions on expression of this kind may be easier to justify than other infringements of s. 2(b)."215 Hate speech is distant from the core values precisely because it not only does little to promote the underlying rationales, but can impede them. Consideration of the extent to which the expression in question fulfills the underlying purpose of freedom of speech, in a limited manner in the course of considering whether the limitation is justifiable, might be an attractive alternative to employing pigeonholes for lesser-protected categories of speech. Another significant difference in approach was reflected in the Court's consideration of intent. The applicant, as well as other critics of these human rights law provisions in general, had argued that the lack of a requirement to demonstrate an intent to foster hatred made the provision too broad. The Court's response does not so much stand in contrast to the American approach to hate speech legislation, as it does to the American approach to discrimination in general. It will be recalled that the Supreme Court of the United States has held that in order to ground a claim of discrimination based on a suspect class such as race, in violation of the equal protection clause of the Fourteenth Amendment, one must establish "invidious intent" to discriminate. Merely establishing that a law or government action has had a "disparate impact" on members of a racial minority (for instance) does not trigger heightened scrutiny. 2 16 The Canadian approach to discrimination under Section 15 of the Charter is quite different, with the Supreme Court having consistently held that laws that have either the intent or the effect of discriminating on one of the prohibited grounds of discrimination will be in violation of the right, subject to the justification analysis. 217 That perspective was similarly reflected in the Court's consideration of hate speech provisions human rights legislation, as it found in both Taylor and Whatcott that because systemic and structural discrimination is more prevalent than intentional forms of discrimination, it is entirely reasonable for the legislation to focus on effect, and not to require proof of discriminatory intent. Indeed, to focus

214. Id. 215. Id. at para. 113 (internal citations and quotes not included). 216. Washingtonv. Davis, 426 U.S. 229 (1976); McClesky v. Kemp, 481 U.S. 279 (1987). 217. Law Society of British Columbia v. Andrews, [1989] 1 S.C.R. 143 (Can.); and Law v. Canada (Minister of Employment and Immigration), [1999] 1 S.C.R. 497 (Can.). Spring 2018] STRIKING THE RIGHT BALANCE 511 on intent, the Court held, would be to defeat the primary goal of antidiscrimination statutes. 218

IV. Finding the Right Balance Having conducted an overview of each of the different approaches to hate speech laws, it now remains to explore how such a comparative analysis may help us in thinking about the scope and substance of an optimal approach to hate speech legislation. An optimal approach would be one that goes some way towards protecting identifiable groups from the very real harm posed by hate speech, while limiting government infringement of the right to freedom of speech to the very minimum necessary to achieve that objective. Others have observed that the approaches to the problem of hate speech in different countries illustrates some of the more fundamental differences in those national systems regarding the conception and priority of rights in different systems. 219 In this sense, the Canadian system reflects the emphasis on equality as a substantive right and the privileging of multiculturalism, while the American system reflects an emphasis on liberty and freedom from government interference over equality and equal protection. 2 20 We might add to this that the Japanese system reflects both a tendency towards the exercise of "authority without power," a weak approach to rights enforcement generally, and most critically, a reluctance to protect the rights of minorities. 221 The complex historical reasons for these different national ideas about the relationship between law and fundamental rights is obviously beyond the scope of this short work-though some of these reasons will be apparent to even the casual reader of history. But these different higher-order national conceptions of rights do, of course, influence and inform the more specific aspects of the different systems as they relate to hate speech laws. Here, I will begin the discussion at only a slightly lower level of abstraction, with the issue of how the different systems view the relationship between hate speech law, the right to equal treatment, and equal protection. This, as discussed briefly at the outset and in passing during the examination of each national approach, is key to the very idea that hate speech laws implicate a tension between the right to equality and the right to freedom of expression.

218. Canada (Human Rights Commission) v. Taylor, [1990] 3 S.C.R. 892, 931-32 (Can.); Whatcott, [2013] 1 S.C.R. 467, para. 126. 219. Moran, supra note 38, at 1425-26; Rosenfeld, supra note 91, at 1523-24. 220. Rosenfeld, supra note 91, at 1523-24. 221. See supra notes 24-28 and accompanying text, and Part I-C. 512 HASTINGS CONSTITUTIONAL LAW QUARTERLY [Vol. 45:3 A. Recognizing the Implication of Equality Rights It will have become obvious from our review of the different systems that there is a considerable difference on this idea that hate speech implicates constitutional equality rights. As we have seen, Canadian law-makers and the Supreme Court of Canada have embraced the idea that hate speech law has the object and purpose of fulfilling the constitutional guarantee of equal protection and equal benefit under the law, while neither the American nor Japanese legislatures and courts have not tended to recognize such a relationship. Indeed, in the context of other tensions between antidiscrimination legislation and First Amendment rights to freedom of speech or freedom of religion, there is a tendency to dismiss as irrelevant any invocation of the Fourteenth Amendment equal protection clause, precisely because the discrimination at issue is among private persons.222 In other words, because the discrimination prohibited by the antidiscrimination law is not within a relationship between the individual and the state, the constitution does not apply. As such, so the argument goes, it is a mistake to suggest that a second constitutional right is implicated at all, far less that it could be in tension with the first right at issue. Given this quite pervasive perspective, it is perhaps necessary to pause and consider in more detail just how fundamental rights might in theory be implicated, and thus stand in tension with other rights, in these sorts of circumstances. It is perhaps best to begin with an example of obvious conflict or tension between constitutional rights. The sharpest and most direct tensions between two constitutional rights arise when the government takes action or enacts law for the purpose of fulfilling or enforcing one right, but in the process, is said to directly limit or infringe a different right. Thus, for instance, when Colorado sought to amend its constitution so as to prohibit the enactment of any law or ordinance extending to gays and lesbians the right not to be discriminated against on the basis of their sexual orientation, it did so on the ostensible grounds that Colorado was thereby seeking to protect the rights of other residents to freedom of religion, freedom of speech, and freedom of association. The Supreme Court, in the case Romer v. Evans,223 struck down the amendment as a violation of the equal protection clause of the Fourteenth Amendment. The Court, in finding that the amendment was motived by animus towards gays and lesbians, clearly questioned the validity of the claim that its true purpose was to protect the right to freedom of religion and

222. See infra notes 227-251 and accompanying text; but see Shannon Gilreath& Arley Ward, Same-Sex Marriage,ReligiousAccommodation, andthe Race Analogy, 41 VT. L. REV. 237 (2016). 223. Romer v. Evans, 517 U.S. 620 (1996). Spring 2018] STRIKING THE RIGHT BALANCE 513 freedom of speech224-but if we were to accept the Colorado argument at face value, it would reflect a kind of direct tension between constitutional rights. The government of Colorado, however, had simply not got the balance right. In seeking to enhance protections for the right to religious freedom-which it was thought could be potentially infringed by the operation of broad antidiscrimination laws-the government of Colorado had unjustifiably violated the separate right of gays and lesbians to equal 225 protection.22 Given that constitutional rights relate primarily to the relationship between the state and the individual, such cases as this constitute a true and direct conflict among competing rights, in a zero-sum game in which one right or the other is going to be more compromised directly by state action, depending on where the line is drawn and how the tension is resolved. And at least from some theoretical perspectives, in which fundamental rights are seen as having similar weight and importance, governments should be seeking to find an equilibrium point at which each right is compromised in equal measure, with each protected to the maximum extent possible without excessively impinging on the competing right, so as to make the necessary impingement on both justifiable. It is quite common, however, for tensions between or among rights to arise in situations in which the conflict is not quite so direct. Rather than cross-cutting government action directly affecting two separate rights of individuals in their relationship with the state, one side of the equation involves the relationship among persons in the private sector. Thus, the government is seen to be directly infringing one rightthrough the enactment of law or policy for the purpose of protecting certain constitutional values from being violated by private actors. By constitutional values, I mean a value that flows from or is informed by the idea that forms the foundation of the right. But while a right gives rise to an enforceable claim, the value does not. The idea that individuals should be free from discrimination on the basis of race forms the foundation of the equal protection clause. It gives rise to a right enforceable against the state, but this constitutionally grounded idea may be characterized as a value that informs government legislation of

224. Id. at 632. 225. Of course, as the Supreme Court suggested in this case, antidiscrimination laws that prohibit businesses from discriminating, in the provision of services or in employment, against people on the basis of sexual orientation, do not actually infringe the right to freedom of religion in a meaningful way. But that issue is heading back to the Supreme Court in Craig v. Masterpiece Cakeshop, Inc., 370 P.3d 272 (Colo. App. 2015), the appeal from which was heard by the Supreme Court in December 2017. 514 HASTINGS CONSTITUTIONAL LAW QUARTERLY [Vol. 45:3 private relations. This is not an idea well developed in American constitutional thought, but is part of constitutional discourse in Canada.226 This idea of limiting a right while trying to give effect to a constitutional value, is reflected in precisely the sort of antidiscrimination laws that were at issue in Romer v. Evans. When such laws are seen to prohibit an individual (or a private corporation) from discriminating against persons on the basis of their sexual orientation, for instance, the law is challenged as violating the individual's right to free exercise of religion. 227 The antidiscrimination law is enacted to prevent private entities from discriminating against people on the basis of race, ethnicity, gender, sexual orientation, and the like, in the context of employment, and access to accommodations and other services. But, this can be viewed as the government seeking to protect a constitutional right, or least the values and interests underlying a constitutional right, from private encroachment. And thus, the enactment and operation of such laws, to the extent that they legitimately impinge on a separate constitutional right, can be said to create a tension between the right to equal protection and that second right, even though the tension is somewhat oblique or indirect. 228 It is this more indirect form of tension that characterizes the dilemma posed by hate speech laws. This is clearly contemplated by the ICCPR and the CERD. As mentioned earlier, the ICCPR explicitly requires countries to prohibit by law "any advocacy of national, racial or religious hatred that constitutes incitement to discrimination, hostility or violence," in furtherance of the separate right to be treated equally and not to be discriminated against on the basis of such characteristics as race, national origin, and religion. Yet, it also requires states to guarantee, respect, and enforce the right to freedom

226. For Supreme Court discussion of the scope and operation of Charter values, albeit in the context of administrative law decision-making, see Dord v. Barreau de Qudbec, [2012] 1 S.C.R. 395 (Can.); and see Lorne Sossin & Mark Friedman, Charter Values and Administrative Justice, 67 S.C. L. REV. 391 (2014). 227. See, e.g., Craig v. Masterpiece Cakeshop, Inc., Colorado Court of Appeals, Case No. 14CA1351, Aug. 12, 2015; cert. grantedJune 26, 2017, argument heard December 5, 2017. 228. To be crystal clear on the difference between these two scenarios, inRomer v. Evans the government was trying to implement a constitutional amendment that would strip antidiscrimination laws of any protections for gays and lesbians, which directly implicated the Equal Protection Clause. On the other hand, the operation of such antidiscrimination laws, in their protection of individuals in private relations from discrimination on the basis of race, sex, sexual orientation, and so forth, seeks to protect the values and interests that the Equal Protection Clause is designed to protect in the public sphere. Where the operation of such law impinges on a separate constitutional right, the tension between that right and equal protection is thus indirect; whereas in Romer v. Evans, the government's proposed constitutional amendment, which would undermine one right in order to enhance the protections of another, created a direct tension between the two rights. Spring 2018] STRIKING THE RIGHT BALANCE 515 of expression. 229 Similarly, the CERD requires state parties to ensure equality before the law in ensuring the right to freedom of expression and opinion, but at the same time obligates state parties to enact laws to prohibit and punish the "dissemination of ideas based on racial superiority or hatred, incitement to racial discrimination . . ." as well as "propaganda activities, which promote and incite racial discrimination." 2 30 The CERD Committee, in one of its General Recommendations on treaty interpretation, quite explicitly articulated the connection between the two rights:

The relationship between proscription of racist hate speech and the flourishing of freedom of expression should be seen as complimentary and not the expression of a zero sum game where the priority given to one necessitates the diminution of the other. The rights to equality and freedom from discrimination, and the right to freedom of expression, should be fully reflected in law, policy, and practice as mutually supportive human rights. 23 1

As we have seen, in the Canadian system hate speech laws have similarly been accepted as having the purpose of implementing within the private sphere the constitutional guarantee of equality and the right not to be discriminated against on the basis of race and similar immutable personal characteristics. Again, such laws target and seek to prevent discrimination perpetrated by private entities, which is discrimination that would not directly implicate constitutional rights or ground constitutional claims. But the government nonetheless views legislative protection against such discrimination as being a fulfillment of the constitutional right to equality. The absence of such legislative protection is moreover seen as constituting a failure to fully realize the equal protection and equal benefit under the law guaranteed by the constitution. Indeed, when Alberta failed to include sexual orientation as a prohibited ground of discrimination in the Provincial antidiscrimination legislation, and a gay teacher challenged the law for its failure to protect him after he was fired because of his sexual orientation, the Supreme Court of Canada found the antidiscrimination law itself to be in violation of the right to equality in Section 15 of the Charter.232 The effort

229. ICCPR, supra note 4, arts. 19, 20, and 26; CERD, supra note 5, arts. 4, 5. 230. CERD, supra note 5, arts. 4, 5. See also Comm. on the Elimination of Racial Discrimination, General Recommendation No. 35, U.N. Doc. CERD/C/GC/35 (2013) [hereinafter CERD Committee Recommendation No. 35]. 231. CERD Committee Recommendation No. 35, supra note 230, atpara. 45. 232. Vriendv. Alberta, [1998] 1 S.C.R. 493 (Can.). 516 HASTINGS CONSTITUTIONAL LAW QUARTERLY [Vol. 45:3 to provide such legislative protection in the form of hate speech laws, however, will necessarily constitute a direct government infringement of the right to freedom of expression, for which the government must provide a powerful justification. And therein lies the tension between constitutional rights, notwithstanding that hate speech addresses the conduct of private entities. Our review of the American approach, however, suggests that hate speech laws are seen first and foremost as a senious threat to freedom of speech. Moreover, and most importantly, the prevailing American view is that even if a failure to prevent hate speech may result in increased levels of discrimination by private entities, private discrimination itself is not a constitutional issue and does not implicate the equal protection clause. That the Fourteenth Amendment equal protection clause should not be understood to apply to private action was established by the Supreme Court in the Civil Rights Cases, soon after the amendment was promulgated.233 The Supreme Court has in other contexts denied that the Bill of Rights requires government to affirmatively protect people from private action.234 And, as others have observed, the American judiciary has generally failed to recognize the extent to which the hate speech cases implicate equal protection rights under the Fourteenth Amendment.235 Even in Beauharnais, the high water mark of hate speech cases, in which the most egregiously racist anti-African American speech was at issue, the Court provided scant indication that equal protection considerations were animating its thinking, either in terms of justifying a limitation on freedom of speech, or informing the purpose and rationale of the law in question.236 Thus, in contrast to the Canadian and international human rights approach, the idea is not well established that constitutional rights might recommend, far less require, government action to provide protection for related values and interests within the private sphere. Governments are of course free to pass laws to provide such protection, but to the extent the resulting laws impinge on another constitutional right, the argument that the law was passed to implement

233. Civil Rights Cases, 109 U.S. 3, 20 (1883). 234. DeShaney v. Winnebago County Dept. of Social Services, 489 U.S. 189 (1989) ("a State's failure to protect and individual against private violence simply does not constitute a violation of the Due Process Clause"). 235. SUNSTEIN, supra note 82, at 185; Rosenfeld, supra note 91, at 90; Moran, supra note 38, at 131-39. 236. Beauharnaisv. Illinois, 343 U.S. 250 (1952). Sunstein makes a similar point. SUNSTEIN, supra note 82, at 185. Spring 2018] STRIKING THE RIGHT BALANCE 517 constitutional protections within the private sphere will not find much purchase as ajustification for the violation of the other constitutional right. 23 7 Yet notwithstanding this mainstream characterization of the American approach, there are some toeholds here and there upon which one could begin to develop a normative argument for the recognition of a relationship between private discrimination and the equal protection clause. For instance, in Romer v. Evans, discussed above, the Supreme Court held that Colorado's effort to deliberately and affirmatively deny a particular class of persons the same protections against private discrimination that were offered to other identifiable groups, constituted a violation of the equal protection clause- even when that group was not defined by a suspect classification. 2 38 More explicitly, the courts have upheld Federal civil rights legislation enacted pursuant to the "enforcement provisions" of the Thirteenth, Fourteenth, and Fifteenth Amendments. The Supreme Court has held that the enforcement provisions permit Congress to enact laws in order to implement and fulfill the purpose of the Reconstruction Amendments, even where such laws tranche on state powers, and interfere with private relations. They vary slightly in wording, but Section 5 of the Fourteenth Amendment provides simply that "[t]he Congress shall have the power to enforce, by appropriate legislation, the provisions of this article."2 39 Thus, in Jones v. Alfred H. Mayer, Co., the Court held that the provisions of the Civil Rights Act of 1866 240-which provided that all citizens of the United States shall have the same rights as are enjoyed by white citizens to, among other things, purchase, lease, sell, hold and convey real property-was a valid law enacted in accordance with the enforcement provision of the Thirteenth Amendment, and could be invoked by an African American who had been discriminated against by private individuals 24 refusing to sell him a house because of his race. 1 Or, to put it more starkly, in order to enforce and implement the Thirteenth Amendment, Congress could enact laws to eliminate racial barriers to acquiring property, notwithstanding that under the Constitution of the United States, property was properly within the jurisdiction of State governments rather than Congress, and the purchase and sale of property was a private transaction. 24 2

237. U.S. v. Morrison, 529 U.S. 598 (2000). 238. Romerv. Evans, 517 U.S. 620, 626-636 (1996). 239. U.S. CONST. amend. XIV, § 5. 240. 42 U.S.C. § 1982 (2006). 241. Jones v. Alfred H. Mayer, Co., 392 U.S. 409, 413 (1968). 242. The Thirteenth Amendment provides that "Neither slavery nor involuntary servitude, except as punishment for crime whereof the party shall have been duly convicted, shall exist within HASTINGS CONSTITUTIONAL LAW QUARTERLY [Vol. 45:3

Similarly, in United States v. Guest,243 two defendants were prosecuted for conspiring to deprive African Americans of the right to use state owned and operated facilities in Georgia, pursuant to a Federal law that made it a crime to conspire to deprive citizens of the free exercise of any right or privilege enjoyed under the Constitution. 244 The accused challenged the law in part because there was no state action involved in depriving the victims of their rights, but the Supreme Court rejected the argument, and several of the concurring Justices held that notwithstanding that the Fourteenth Amendment did not apply to private action, the enforcement provision empowers Congress to "enact laws punishing all conspiracies-with or without state action-that interfere with Fourteenth Amendment Rights."245 This must be tempered, however, by acknowledging that the Court has more recently foreclosed most arguments that Congress can enact national antidiscrimination law under the authority of Section 5 of the Fourteenth Amendment. In US. v. Morrison,24 6 which involved Federal legislation providing a civil cause of action for victims of gender-based violence, a slim majority of the Court questioned the validity of the concurring opinions in Guest, and denied that Congress could, under authority of the enforcement provision of the Fourteenth Amendment, enact laws directed exclusively at the discriminatory action of private persons.247 The dissent, relying on such cases as City of Boerne v. Flores,248 argued that the Court had in the past upheld Federal remedial legislation, enacted under this same authority, that applied to private actors for conduct that would not itself be unconstitutional-and that doing so could influence states to provide stronger protections.249 Morrison may have made more difficult any arguments that Congress could invoke the enforcement provision of the Fourteenth Amendment to enact antidiscrimination law directed at private action (though on issues of

the United States, or any place subject to their jurisdiction." U.S. CONST. amend. XIII, § 1. The Court reasoned that the enforcement provision gave Congress authority to pass laws necessary to abolish all "badges and incidents of slavery," and that the provision of the Civil Rights Act of 1868 did just that. Jones, 392 U.S. at 439-440. 243. United States v. Guest, 383 U.S. 745, 761-62 (1966) (Clark, J., Black, J., and Fortas, J. concurring). 244. 18 U.S.C. § 241 (1964). 245. Guest, 383 U.S. at762. 246. United States v. Morrison, 529 U.S. 598 (2000). 247. Id. at 620-22. 248. City of Boerne v. Flores, 521 U.S. 507 (1997). 249. Morrison, 529 U.S. at 665-66 (2000) (Breyer, J., dissenting, joined by Souter, J., Ginsburg, J., and Stevens, J.). Spring 2018] STRIKING THE RIGHT BALANCE 519 race, the enforcement provision of the Thirteenth Amendment remains more available). And yet, there are other such Federal antidiscrimination laws, arguably animated by the same values that form the foundation of the equal protection clause of the Fourteenth Amendment, and which, moreover, involve subject matter limits on expression. Antidiscrimination provisions that form part of Title VII of the Civil Rights Act of 1964,250 for instance, include prohibitions on discrimination comprising of conduct, including speech, that contributes to the creation of "intimidating, hostile, or offensive 2 5 working environment." 1 In a case involving claims of sexual harassment, the Supreme Court endorsed and upheld this understanding of discrimination in the legislation.252 The similarity between expression that causes a hostile work environment and hate speech is not hard to see, and indeed I am arguing for a hate speech law that would be far more narrowly drawn. Yet, while the Title VII provisions are surely drawing upon the constitutional values animating the equal protection clause, they were not explicitly enacted to fulfill the Fourteenth Amendment. No First Amendment challenge to the workplace harassment provisions have yet been decided by the Supreme Court.253 In sum, while there is little recognition of any relationship between the Equal Protection Clause and legislation seeking to prevent private discrimination caused by hate speech, there is some basis elsewhere in American jurisprudence for developing an argument for recognizing such a relationship-or at least the recognition that hate speech laws are animated by the same constitutional values as the Equal Protection Clause. Similar arguments can be made about Japan. As indicated in Part II above, there has been no indication that the Diet recognizes that antidiscrimination laws in general, or the recent hate speech legislation in particular, has either the purpose or effect of fulfilling the right to equality in Article 14 of the Constitution. But the courts have in the past interpreted legislation in a manner that was quite explicitly informed by constitutional rights, and indeed by the right to equality in Article 14. Thus, in the famous Nissan Motors case, involving discrimination against women in the workplace, the Supreme Court endorsed the manner in which the lower courts had employed the constitutional values of the equality right in Article 14 to inform and give substance to statutory provisions governing the private

250. 42 U.S.C. §. 2000(e) etseq. 251. Meritor Bankv. Vinson, 477 U.S. 57, 65-66 (1986). 252. Id. My thanks to Bill Rich for directing my attention to this line of argument. 253. For analysis of the issue, see e.g., Eugene Volokh, Thinking AheadAbout Freedom ofSpeech and Hostile WorkEnvironment Harassment, 17 BERKELEY J. EMP. & LAB. L. 305 (1996). 520 HASTINGS CONSTITUTIONAL LAW QUARTERLY [Vol. 45:3 relationship between employers and employees. 254 Moreover, in the Kyoto Korean School case discussed in Part I above, the Kyoto District Court's holding that the defamatory speech came within the scope of the definition of discrimination in the CERD, was only one step removed from recognizing that limiting such speech is for the purpose of enforcing a right to equality by preventing discrimination.25 5 Thus, even more so than for the United States, there is a basis for suggesting that legislatures and courts can recognize the role of constitutional rights, or constitutional values, in informing the purpose and scope of legislation governing private relations. There is, therefore, a starting point in both the United States and Japan for arguing that hate speech laws could be understood as fulfilling the purpose of a constitutional right to equality, or at a minimum as being informed by the constitutional values underlying that right. But there is also in both countries evidence that neither the legislatures nor the courts take sufficiently seriously the nature and extent of the harm that can be caused by hate speech. There is a wealth of social science literature on such harm, and it has been decades since that evidence has been brought into the mainstream of legal scholarship.256 As reviewed above, the Supreme Court of Canada recognized and accepted that evidence of harm almost thirty years ago. 25 7 What is more, the gravity of that harm informed the Court's analysis of whether the infringement of the right to freedom of expression caused by narrow hate speech laws could be justified. 2 58 As Mayo Moran has examined in some detail, American judges have tended to minimize and even trivialize the nature and extent of such harm, reducing the concern to "hurt feelings" and the like, and suggesting that the effect of "mere words" is transient.259 In contrast, the American courts have readily justified content-based, and even viewpoint-based, limits on speech when the harm that the law is intended to avoid is viewed as sufficiently serious or grave, notwithstanding the august position of the First Amendment right to freedom of speech in the pantheon of rights and liberties. The fact that the limits imposed by hate

254. Supreme Court Judgment, (Mar. 24, 1981), 35 Minshii 2 300 (Nissan Motors case) (holding that "a lower compulsory retirement age for women than for men constitutes discrimination against women based solely on their gender and is irrational discrimination invalid under Article 90 of the Civil Code . . . Article 1-2). For an English translation of this case, under the title Nissan Motors, Inc. v. Nakamoto, see BEER & ITOH, supra note 67, at 179-81. 255. See supra notes 18-23 and accompanying text. 256. For examples, see supra note 38 and 174-81. 257. R. v. Keegstra, [1990] 3 S.C.R. 697 (Can.). 258. See supra text accompanying notes 174-181, and 186-187. 259. Moran, supra note 38; see also SUNSTEIN, supra note 82, at 186; and see Rosenfeld, supra note 91, at 1559-61. Spring 2018] STRIKING THE RIGHT BALANCE 521 speech law are neverjustified suggests that the harm posed is not sufficiently recognized; and the corollary is that if the harm was fully appreciated, narrowly tailored hate speech law could be upheld. As Cass Sunstein sums it up, viewpoint-based limitation will be permitted "when there is no serious risk of illegitimate government motivation, when low-value or unprotected speech is at issue, when the skewing effects on the system is minimal, and government is able to make apowerful showing ofharm."260 Laws designed to prevent the serious harm caused by hate speech, and in particular fulfill the constitutional guarantee against unjust discrimination, satisfy this test. I want to suggest that it is crucially important to recognize that hate speech laws can and should implicate equality rights-that they ought to be crafted with a view to fulfilling the promise of constitutional equal protection rights. And that as such, they thus trigger a tension between two fundamental constitutional rights. This is not simply because I take the harm of hate speech seriously and want to develop protections against the discrimination that it can spawn-though that is certainly true too. Rather, it is because this recognition gives rise to a compelling logic, an almost naturally developed and entirely coherent argument in favor of the justification of narrowly tailored hate speech laws. In contrast to laws that grope for an appropriate purpose and objective, and for which a constitutional justification is hard to find, this argument is naturally shaped by an understanding of hate speech laws as furthering the values and protecting the interests that are enshrined in the constitutional equality rights provisions of most constitutional democracies. In the absence of any underlying constitutional imperative, for instance, the American approach has tended to gravitate towards imminent violence and disruption of the peace as the limiting principle for laws that might justifiably limit expression that is in the realm of hate speech. The underlying concern driving this approach is that to look for justification that is less significant than the prevention of violence, is to begin down a slippery slope upon which there will be no principle of sufficient gravity to find purchase. But the Canadian approach suggests just such a principle-one informed by a constitutionally grounded right to equality and not to be discriminated against. It is the prohibition of expression that is likely to foster hatred, contempt, and vilification of the most serious kind, which is likely to lead to harmful discrimination and persecution, and is far from the core of the right to free expression. The concern is not imminent violence or disturbance of the peace, but the serious harm that hate speech can cause to not only the targets of hatred and the victims of the resulting discrimination and persecution, but to the democratic values and the rule of law itself. It is

260. SUNSTEIN, supra note 82, at 177. 522 HASTINGS CONSTITUTIONAL LAW QUARTERLY [Vol. 45:3 ultimately a concern for the protection of the constitutionally grounded right to equal protection and equal benefit of the law. This provides a powerful constitutionally based explanation for the object and purpose of hate speech laws. What is more, it also provides the basis for ready-made and internally coherent limitations on the scope of the laws' impact on free expression. As such, with a compelling objective and narrowly tailored limits, it thus lends itself to the possibility of justification under most judicial doctrines applicable to the right to freedom of expression-even, arguably, the strict scrutiny standard of review in the United States. In the remaining sections, I turn to explore in more detail how the logic of this argument plays out in the language of specific hate speech provisions, and in the judicial doctrine employed to assess the constitutionality of such laws.

B. Natural Limits of Hate Speech Laws As was illustrated by our review hate speech related legislation above, there has been a considerable range of stated legislative objectives, and underlying concerns driving such objectives, which in turn has translated into a considerable variance in the nature and scope of expression that was prohibited. It should be self-evident that differences in object and purpose will lead to differences in scope and substance of the limitations of such laws. Let us then begin by considering in more detail the object and purpose of the laws we have reviewed. It will be recalled that in the American cases, with laws that were formulated (or interpreted as having been formulated) to limit fighting words, the concern was that the prohibited speech could provoke a violent reaction on the part of the subject and target of the speech, violence that would be directed towards the speaker, and so disturb the peace.26' In the case of laws that sought to prohibit incitement to violent or unlawful conduct, the concern was just that--that the prohibited speech could provoke the target audience to commit violent or criminal acts, directed at someone other than the speaker, typically but not always the subject of the speech.262 In the case of both fighting words and incitement, the overarching object and purpose was to prevent imminent violence and disturbances of the peace, not necessarily to prevent harm to the listener, or even the subjects of hatred or contempt in the speech. It will be recalled that in Brandenburg, while the speech vilified African Americans, the prosecution was actually based on incitement to violence against the government.263 It is true that in the case

261. See supra Part II-B. 262. See supra Part II-B. 263. Brandenburgv. Ohio, 395 U.S. 444 (1969). Spring 2018] STRIKING THE RIGHT BALANCE 523 of laws seeking to prevent a species of true threats, as in Virginia v. Black, the concern is the potential harm to the persons who were the target of the threat, in terms of fear and distress the threat would trigger. But even here, the element of imminent violence is again lurking just below the surface.264 Turning to Japan, we saw that the recent ground-breaking litigation was grounded in laws that prohibited criminal and civil libel, insult, and obstruction of commerce, all of which required that there be specific victims who were the target of the expression and conduct in question. None of these laws have the object and purpose of preventing the fostering of hate or preventing discrimination, but were merely available for the prosecution of a very specific instance of hate speech.265 As for the new hate speech law, the object and purpose is indeed to reduce unfair and discriminatory speech, though it is without any reference to the constitutional right to equality, and it is for the limited protection of a very narrowly defined category of persons that constitute only a small subset of all the identifiable groups that are protected by Article 14 of the Constitution. 26 6 In contrast to the U.S. and Japan, however, Canadian hate speech legislation, in terms of both the criminal law and the civil human rights law provisions, the overarching object and purpose was to prevent the fostering of hatred that could contribute to increased discrimination and persecution, as explicitly informed by the constitutional equality right. The overriding concern was the serious harm to the identifiable group targeted by the hate speech, both directly in terms of creating in members of the group a sense of alienation and loss of status within the community, and through the increased discrimination against members of the group by the majority, who are often the true target audience of the hate speech. Imminent violence or lawlessness is not the concern--though, it will be recalled, other Canadian laws are designed to address that separate threat. 267 These differences in object and purpose, and the overarching concerns that inform those objectives, in turn affect the scope and subject matter of the laws. Because the Canadian legislation is explicitly informed by the constitutional equality right, with the purpose of preventing discrimination, it is only natural that the parameters of the legislation be defined by the prohibited grounds of discrimination in the constitutional equality right. Thus, the Criminal Code provision prohibits the promotion of hatred against

264. See supra Part II-C. 265. See supra Part I-A. 266. See supra Part I-B. 267. Criminal Code of Canada, R.S.C. 1985, c C-46, (as amended), s. 319(1). 524 HASTINGS CONSTITUTIONAL LAW QUARTERLY [Vol. 45:3 any "identifiable group," a concept that is defined in a manner that exactly tracks the prohibited grounds of discrimination in Section 15 of the Charter, namely color, race, religion, national or ethnic origin, age, sex, sexual orientation, gender identity, or disability.26 8 The Saskatchewan Human Rights Code provision considered above similarly prohibited any representation that would expose to hatred any person or class of persons on the basis of the prohibited grounds in the Act, which again track the prohibited grounds of discrimination in Section 15 of the Charter.26 9 The constitutional right thus provides the natural limits-the prohibited grounds of discrimination-with which to define the narrow scope of the law crafted to prevent discrimination. These limits are further narrowed by relying on the concept of hatred, which as was examined above, is intended to represent the most extreme and intense feelings of detestation and vilification. And so the law seeks to prohibit expression that takes an extreme form designed to foster and incite such intense emotions, through denigration, de- legitimization, and dehumanization of the members of the identifiable group, which it is understood will in turn lead to increased discrimination and persecution, on the basis of shared personal characteristics that form the prohibited grounds of discrimination in Section 15 of the Charter.270 Limited in this manner, the Criminal Code of Canada hate speech provision is actually much narrower than most of the American laws that were the subject of constitutional challenges discussed above. Consider the ordinance in R.A. V v. St. Paul, which prohibited the employment of symbols or objects such as swastikas or burning crosses, "which one knows or has reasonable grounds to know arouses anger, alarm or resentment in others on 27 the basis of race, color, creed, religion, or gender." 1 Here the law does in fact reflect several of the grounds of discrimination that are proscribed by the Constitution, but the other limiting element, the emotional response of listeners, including mere "resentment," is far broader than the concept of "hatred" employed in the Canadian law. The Supreme Court of Canada made quite clear that the prevention of such emotions as anger, resentment, or other forms of offense, is not remotely sufficient to justify an infringement on freedom of expression. The language of the law at issue in R.A. V v. St. Paul would likely be struck down in Canada for this reason, just as the clause

268. Id. at s. 319-318. It will be noted that there are characteristics, such as sexual orientation, included in s. 319 that are not explicitly provided for in s. 15 of the Charter, but these have been held to be "analogous grounds" by the Supreme Court, and are understood to be within the scope of s. 15. 269. The Saskatchewan Human Rights Code, S.S. 1979, c S-24.1, at s. 14(1)(b) (Can.). 270. See supra text accompanying notes 200-205. 271. R.A.V. v. City of St. Paul, 505 U.S. 377 (1992). Spring 2018] STRIKING THE RIGHT BALANCE 525 of the Saskatchewan Human Rights Code, that purported to prohibit speech that "ridicules, belittles or otherwise affronts the dignity of any person" was struck down.272 It will be recalled, however, that there was considerable difference between the Criminal Code of Canada provision, and the Saskatchewan Human Rights Code provision, in that the latter did not require any element of intent, and did not afford any explicit defenses. The Criminal Code provision prohibits the willful promotion of hatred, which establishes a significant specific-intent requirement. From an American perspective, this might seem to be a necessary limitation, and any law that simply prohibits speech that may be likely to have the effect of fostering hatred and contributing to discrimination, whether intended or not, may go too far. This is particularly so in light of the fact that government action that cannot be shown to have had an invidious discriminatory intent to discriminate, is not subject to heightened scrutiny in equal protection claims under the Fourteenth Amendment.273 Thus, for purposes of drawing potentially viable examples of model legislation from our comparative review, the significantly narrower Criminal Code provision is likely to be the more acceptable as a basis for developing a new American approach to hate speech laws. It should also be noted that the prohibited grounds of discrimination in both Canada and Japan are far more extensive than in the United States. The list of "suspect" and "quasi-suspect" classes that attract heightened scrutiny in the United States is relatively short, comprising of race, national origin, sex, and legitimacy (birth status); while discrimination on the basis of religion is prohibited under the First Amendment. Candidate suspect classes, such as sexual orientation, remain to be conclusively decided- while they have clearly attracted something higher than the typical rational- basis review, they are not yet recognized as suspect classes. 274 So the "natural limits" of hate speech as informed and defined by the constitutional right to equal protection in the American context might also be necessarily narrower than elsewhere.

C. Implications for Doctrine and Justification Hate speech law that is limited in accordance with this logic, also tends to more naturally comply with the doctrine for justifying limitations on freedom of expression. As our review of the Canadian cases reflects, the

272. Whatcott v. Saskatchewan Human Rights Tribunal, [2013] 1 S.C.R. 467, paras. 89, 111 (Can.). 273. Washingtonv. Davis, 426 U.S. 229, 247-248 (1976); see supra text accompanying notes 216. 274. See, e.g. Romerv. Evans, 517 U.S. 620 (1996). 526 HASTINGS CONSTITUTIONAL LAW QUARTERLY [Vol. 45:3 recognition that the object and purpose of the hate speech law is for the fulfillment of the constitutional protections enshrined in the equality right provision of the Charter, together with a fuller appreciation of the serious harm that hate speech can cause, provides the most powerful "compelling state interest" for limiting the right to freedom of expression. Indeed, for those who think that an infringement of a fundamental constitutional right can only be legitimately justified by reference to the protection or enforcement of some other fundamental constitutional right,27 5 this objective is the very best available. Similarly, once the law is narrowly drawn in accordance with the natural limits provided by the prohibited grounds of discrimination, together with the concept of hatred as described above, it is not difficult to characterize the law as being carefully tailored and minimally impairing the right to freedom of expression. This carefully tailored law will only proscribe the most extreme forms of expression that are intended to foster hatred and are likely to lead to increased discrimination and persecution. One likely objection is this rosy characterization is that no matter how you slice it, hate speech law along these lines is necessarily viewpoint-based, and therefore is going to run afoul of the American doctrine as reflected in R.A. V v. City of St. Paul. There are several different possible responses to this objection. First, I would argue that even if we concede that the law is viewpoint-based, it can survive strict scrutiny. Second, it can be argued that the prohibitions along the lines of the Criminal Code of Canada provision do not actually constitute a viewpoint-based limitation. Third, I would argue that another way to get around the specific problem posed by viewpoint- based limits within a proscribed category of speech, is to establish a new category of lesser-protected speech encompassing just this kind of hate speech, such that there is then no viewpoint discrimination within or among the proscribed or lesser protected category of speech, as was the case in R.A. V v. City ofSt. Paul. Let me examine these three points in more detail. To begin with the concession that the proposed form of hate speech law is necessarily a viewpoint-based limitation on speech, I would argue that if the justification analysis is conducted in a rigorous and meaningful way, such a law can nonetheless survive strict scrutiny. R.A. V v. City of St. Paul confirmed that within American constitutional law, the presumption is that viewpoint-based limits are impermissible, but it is only a presumption, and such laws are still subject to justification under a strict scrutiny review. 27 6 Part of the criticism of the American cases in this area is that the Court does

275. See, e.g., STUN SMITS, RESOLVING CONFLICTS BETWEEN HUMAN RIGHTS: THE JUDGE'S DILEMMA, Chap. 4 (2017); Stephen Gottlieb, Compelling Governmental Interests:An Essentialbut Unanalyzed Term in ConstitutionalAdjudication, 68 B.U. L. REV. 917 (1988). 276. For discussion of this rebuttable presumption, see SUNSTEIN, supra note 82, at 184-192. Spring 2018] STRIKING THE RIGHT BALANCE 527 not tend to engage seriously in the application of strict scrutiny. In R.A. V v. City of St. Paul, for instance, Justice Scalia accepted with little discussion that the law sought to advance a compelling state interest, but then he summarily concluded that the law was not the least restrictive alternative, without any analysis of what the other alternatives might have been, far less how effective they would be. 277 But in the judicial review of the proposed form of hate speech law, where the compelling state interest is to fulfill the protections guaranteed by another constitutional right, and where the law's viewpoint-based limits are in fact informed by and reflect the "viewpoint" of that other constitutional right, there is reason to believe that it could survive strict scrutiny. The primary reason for the concern around viewpoint-based limits, is that we should be very suspicious of government motives in seeking to stifle or suppress any particular viewpoint. But this suspicion should surely be allayed if the viewpoint is directly informed by the Constitution itself, or to put it another way, the viewpoint represents a constitutional value. Survival of strict scrutiny is all the more plausible where the Court, in contemplating the nature of the compelling state interest, also takes seriously the evidence of significant harm that the law is designed to prevent, and considers that harm in contrast to the very limited manner in which the law impinges on the right to freedom of speech. Justice Stevens in particular has on a number of occasions encouraged the Court to incorporate such a proportionality test into the strict scrutiny analysis. 278 And even though the Court has never formally adopted such a proportionality test, and tends not to rigorously apply strict scrutiny in viewpoint-based cases, as discussed earlier, there are certainly other viewpoint-based limits on freedom of speech that have survived judicial review. 2 79 Examples include the ban on advertising in favor of gambling at casinos,280 limitations on speaking against the effects of 28 unionization prior to a union election. 1 As mentioned previously, such viewpoint-based discrimination tends to be permissible when there is minimal risk that the government's motive in imposing the limitation is illegitimate or suspect, the skewing effects are minimal, and the limit is for the purpose of preventing a significant demonstrable harm.282 The second argument would be that the proposed hate speech law does not actually constitute a viewpoint-based limitation at all. This argument

277. R.A.V. v. City of St. Paul, 505 U.S. 377, 395 (1992). 278. See, e.g., Clebume v. Cleburne Living Center, 473 U.S. 432, 451-53 (1985). 279. SUNSTEIN, supra note 82, at 174-78. 280. Posadasv. Louis Co. of Puerto Rico, 478 U.S. 328 (1986). 281. NLRB v. Gissel Packing Co., 395 U.S. 575, 618-19 (1969). 282. SUNSTEIN, supra note 82, at 176. 5289 HASTINGS CONSTITUTIONAL LAW QUARTERLY [Vol. 45:3 could actually take a few different paths. One tempting avenue is to suggest that the form of hate speech law in question is really a harm-based limitation rather than a viewpoint-based limitation. That is, the limit is aimed at preventing a particular form of harm, rather than preventing a particular viewpoint or suppressing a particular side of any argument. This may seem apt here, given the extent to which the proposed form of hate speech law is very much focused on the nature of the harm that the fostering of hatred and discrimination can cause. But Justice Elena Kagan, in an article written when she was still a professor at the University of Chicago, explained that even if one characterizes hate speech laws as harm-based rather than viewpoint-based, they still raise difficult questions.283 Indeed, in one sense, the characterization as harm-based simply disguises the extent to which the regulation is viewpoint-based. Even if the purpose of limiting speech is to prevent the perceived harm that it causes, rather than the ideas as such, the law is nonetheless likely to be based on a particular view of the speech, and thus the law is likely to constitute a viewpoint-based (or at a minimum, content-based) limitation.284 The distinction thus tends to collapse in on itself, and at some point, becomes meaningless-all limitations can be said to be intended, at some level of abstraction, to prevent some harm. Cass Sunstein and others have made similar observations.285 Sunstein, however, also points to a second possible line of argument. He has argued that the Court's formulation of presumptions against viewpoint-based limits in R.A. V v. City ofSt. Paul was appropriate, but that it misidentified the limits in question as being viewpoint-based. Because the law in question prohibited speech that would arouse anger or resentment on the basis of race, for instance, it covered both anti-white and anti-black sentiment (among other possible racist positions), and did not take a position as between these two "viewpoints." It was true that the law singled out a subset of the category of fighting words, and thus was a content-based limitation, but it did not "single out a particular message for prohibition ... [i]t has regulated on the basis of subjects for discussion, not on the basis of viewpoint." 286 To be sure, under the American approach content-based limits are still subject to strict scrutiny, but are viewed with less suspicion, as less presumptively unconstitutional, than are viewpoint-based limitations. And thus, if the law is characterized as a content-based limitation, where the

283. Kagan, supra note 148, at 878-79. 284. Id. at 880-81. 285. SUNSTEIN, supra note 82, at 174-76. 286. SUNSTEIN, supra note 82, at 188. Spring 2018] STRIKING THE RIGHT BALANCE 529 content is defined as representing the values forming foundation of another constitutional right, it should be that much more amenable to justification along the lines laid out above. 287 The third and final argument in response to the objection on viewpoint- based limitation grounds, is that if the U.S. courts were to recognize and establish a new category of lesser-protected speech to encompass hate speech laws, then such laws would not constitute discriminatory viewpoint-based limits within the lesser protected category of speech. It will be recalled that in R.A. V v. City ofSt. Paul Justice Scalia stated that it had been open to the City of St. Paul to proscribe all fighting words-that is, all forms of speech within the proscribable category-it just could not carve out and discriminate against particular viewpoints within the category. If a new category of lesser-protected speech was defined as covering speech intended to foster hatred against identifiable groups on the basis of shared personal characteristics that constitute suspect or quasi-suspect classes under the equal protection clause, then hate speech laws similarly defined would not be making any distinctions among categories within the proscribed speech. This argument for creating a new category of lesser-protected speech within the American constitutional system is bolstered again by the recognition of the relationship between hate speech law and the constitutional right to equal protection, and thus the tension between the right to freedom of speech and the right to equal protection. In the same manner that the content, scope, and limits of hate speech law tend to be naturally defined by the logic that flows from recognizing this relationship, so too the relevant doctrine for purposes of judicial review of hate speech law should be naturally informed and shaped by the implication of two constitutional rights in tension. As our review of the cases has illustrated, hate speech does not "fit" or conform to any of the categories of "proscribable" or lesser- protected speech that have been invoked in cases relating to hate speech. The concerns that underlie the rationale for those categories are quite different. And yet, as we have seen, the narrowly defined speech that would be subject to the proposed hate speech laws are far from the core rationales for freedom of speech. For while "political" in one sense, both the form and purpose of such speech--to foster hatred against an identifiable group and to isolate and alienate them within society, and indeed to drive them out of the marketplace of ideas-is not consistent with underlying rationale for protecting political expression in a democracy. So long as the American approach continues to embrace the use of discrete categories of lesser-protected speech as part of

287. This argument is made in more detail in Little, Hating Hate Speech, supra note 133. 288. R.A.V. v. City of St. Paul, 505 U.S. 377, 388-91 (1992). 530 HASTINGS CONSTITUTIONAL LAW QUARTERLY [Vol. 45:3 the doctrine for justifying limits on expression, the foregoing analysis provides a powerful rationale for establishing a new narrowly defined category of hate speech as one those categories of lesser-protected speech.

Conclusion: Striking the Right Balance What does the foregoing examination of different approaches tell us about the "right balance" and how to strike it? I do not think that any system has found the perfect equilibrium, in terms of where to draw the line between what level of hate speech is permissible and that which is not. In that sense, no one has got the balance "right." But in terms of at least identifying exactly what should be balanced, I would suggest that the Canadian approach gets closer to getting the issue right, and points us in the right direction. In particular, it is the only one of the three systems that actually corresponds to the international human rights law view, that the balance to be struck is one between two separate fundamental rights, as opposed to being a balance between freedom of speech and a mere piece of social-policy based legislation. In the detail of its approach the Canadian system may not have found the perfect equilibrium in resolving this tension between the two competing rights, but it is at least trying to balance the right things. In this sense, it is striking the right balance. This is because, as I have argued, there is a powerful logic and internal coherence that flows from an understanding that hate speech law should be intended to fulfill the promise of constitutional equality rights, thereby creating a tension between two constitutional rights. Hate speech laws enacted with this purpose get at the true harm that we should be concerned about, which is the fostering of hatred against identifiable groups, which causes direct psychological harm to the members of such groups, results in an increase in discrimination and even persecution against the members of such groups, and leads to the more amorphous injury to democratic principles, and even to the right to freedom of expression itself. The purpose of such laws, which are intended to fulfill a constitutional right to equality and protection from discrimination, and to prevent the serious harms that this form of hate speech poses, constitutes a compelling state interest. That compelling state interest should be able to satisfy the first element of any justification analysis in the judicial review of the consequent infringement on freedom of expression. In the context of constitutional judicial review of such a law, the law's explicit objective of furthering the fulfillment of a constitutional guarantee cannot help but be a compelling state interest. But in addition to this, laws so enacted will also naturally constitute narrow but reasonable limits on freedom of expression when subjected to judicial review. I say naturally, because of the manner in which the purpose and Spring 2018] STRIKING THE RIGHT BALANCE 531 function of the law is informed by constitutional values underlying the right to equality-and thus the limits of the law will be shaped by the prohibited grounds of discrimination. While other hate speech laws flounder in search of sufficiently important purposes and limiting principles, such as imminent violence or other disturbances of the peace, hate speech laws with this explicit constitutional objective have tailor-made limitations that are virtually logically required-that the law prohibit speech that is intended to and will likely increase the very kind of discrimination that is prohibited by the constitution itself Now, it is true that the constitution does not provide the standard for where to draw the line on such likelihood of certain speech causing discrimination. That is to say, for instance, where to draw the line on the spectrum between the articulation of policy rationales for restricting immigration from certain countries at one end, to vitriolic calls for the "elimination" of certain ethnic groups, employing the most vile and dehumanizing terminology at the other end. But there is much to commend the Canadian approach of using the concept of hatred and restricting the law to the prohibition of such forms of expression that are likely to foster the most extreme forms of hatred, vilification, and contempt. This concept, while not contained in the constitution of any of the systems under study, is suggested by the evidence of the serious harm caused by precisely such extreme forms of hate speech, and by its relationship to the kind of discrimination the constitution does prohibit. A law prohibiting such hate speech cannot help but impinge upon the right to freedom of expression. It is necessarily both a content-based and subject-matter limitation, and from some perspectives may be characterized as a viewpoint-based limitation on speech. But once the hate speech law has the object and purpose of fulfilling the constitutional right to equality, its form is necessarily tailored by principles of non-discrimination in that right, and the harm that it addresses is fully recognized and internalized into the analysis, then the hate speech law's very limited infringement of the right to freedom of expression triggers a tension between two fundamental constitutional rights-or, at a minimum, between the constitutional values that are the foundation of such rights. This understanding must inform and shape how courts will approach the judicial review of such hate speech laws, and assess the extent to which their infringement of freedom of expression can be reasonably justified. Finding the appropriate balance between them will remain difficult, but at least then we will be considering the right balance. This is certainly possible in the context of Japan, but for reasons I have explained above, there is even room to suggest that it is feasible in the United States. 532 HASTINGS CONSTITUTIONAL LAW QUARTERLY [Vol. 45:3